diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,194621 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.4696067612483728,
+  "eval_steps": 500,
+  "global_step": 27800,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4063446521759033,
+      "learning_rate": 5e-06,
+      "loss": 1.675,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.617262125015259,
+      "learning_rate": 1e-05,
+      "loss": 1.6804,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4406676292419434,
+      "learning_rate": 1.5e-05,
+      "loss": 1.7554,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5562903881073,
+      "learning_rate": 2e-05,
+      "loss": 1.676,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.818110942840576,
+      "learning_rate": 2.5e-05,
+      "loss": 1.6952,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2935879230499268,
+      "learning_rate": 3e-05,
+      "loss": 1.8583,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9137771129608154,
+      "learning_rate": 3.5e-05,
+      "loss": 1.6581,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4779632091522217,
+      "learning_rate": 4e-05,
+      "loss": 2.5016,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4272797107696533,
+      "learning_rate": 4.5e-05,
+      "loss": 1.6304,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.741781234741211,
+      "learning_rate": 5e-05,
+      "loss": 1.6665,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5775020122528076,
+      "learning_rate": 4.9999997356733646e-05,
+      "loss": 1.8808,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8330698013305664,
+      "learning_rate": 4.999999471346728e-05,
+      "loss": 1.7241,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7938928604125977,
+      "learning_rate": 4.999999207020093e-05,
+      "loss": 1.4552,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0222482681274414,
+      "learning_rate": 4.999998942693457e-05,
+      "loss": 1.7172,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.283169984817505,
+      "learning_rate": 4.999998678366821e-05,
+      "loss": 1.6463,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6803457736968994,
+      "learning_rate": 4.999998414040185e-05,
+      "loss": 1.6246,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7833316326141357,
+      "learning_rate": 4.999998149713549e-05,
+      "loss": 1.6513,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.45908522605896,
+      "learning_rate": 4.9999978853869136e-05,
+      "loss": 1.675,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.175886154174805,
+      "learning_rate": 4.999997621060278e-05,
+      "loss": 1.7342,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1730477809906006,
+      "learning_rate": 4.999997356733642e-05,
+      "loss": 1.6891,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.117173433303833,
+      "learning_rate": 4.999997092407006e-05,
+      "loss": 1.523,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.593247413635254,
+      "learning_rate": 4.99999682808037e-05,
+      "loss": 1.6781,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.748931884765625,
+      "learning_rate": 4.9999965637537345e-05,
+      "loss": 1.5772,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6198129653930664,
+      "learning_rate": 4.999996299427099e-05,
+      "loss": 2.1415,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0461316108703613,
+      "learning_rate": 4.9999960351004625e-05,
+      "loss": 1.6863,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.830333948135376,
+      "learning_rate": 4.999995770773827e-05,
+      "loss": 1.6802,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.431565046310425,
+      "learning_rate": 4.999995506447191e-05,
+      "loss": 1.5712,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2176668643951416,
+      "learning_rate": 4.9999952421205555e-05,
+      "loss": 1.5472,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3476085662841797,
+      "learning_rate": 4.99999497779392e-05,
+      "loss": 1.4523,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5906386375427246,
+      "learning_rate": 4.9999947134672835e-05,
+      "loss": 1.499,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.451503276824951,
+      "learning_rate": 4.999994449140648e-05,
+      "loss": 1.515,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7432265281677246,
+      "learning_rate": 4.999994184814012e-05,
+      "loss": 1.7069,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9150450229644775,
+      "learning_rate": 4.9999939204873765e-05,
+      "loss": 1.5093,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.559255599975586,
+      "learning_rate": 4.99999365616074e-05,
+      "loss": 1.5273,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6460278034210205,
+      "learning_rate": 4.9999933918341045e-05,
+      "loss": 1.7335,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9723217487335205,
+      "learning_rate": 4.999993127507469e-05,
+      "loss": 1.6775,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.250629186630249,
+      "learning_rate": 4.999992863180833e-05,
+      "loss": 1.5019,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3502020835876465,
+      "learning_rate": 4.9999925988541975e-05,
+      "loss": 1.6545,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.599980354309082,
+      "learning_rate": 4.999992334527561e-05,
+      "loss": 1.7486,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.356165647506714,
+      "learning_rate": 4.9999920702009255e-05,
+      "loss": 1.5801,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9439274072647095,
+      "learning_rate": 4.999991805874289e-05,
+      "loss": 1.5423,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.687696695327759,
+      "learning_rate": 4.999991541547654e-05,
+      "loss": 1.7454,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4825809001922607,
+      "learning_rate": 4.999991277221018e-05,
+      "loss": 1.5603,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.514963388442993,
+      "learning_rate": 4.999991012894382e-05,
+      "loss": 1.5529,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8896781206130981,
+      "learning_rate": 4.9999907485677465e-05,
+      "loss": 1.4907,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.633730888366699,
+      "learning_rate": 4.999990484241111e-05,
+      "loss": 1.732,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0269463062286377,
+      "learning_rate": 4.999990219914475e-05,
+      "loss": 1.3297,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9929157495498657,
+      "learning_rate": 4.999989955587839e-05,
+      "loss": 1.1349,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2316508293151855,
+      "learning_rate": 4.999989691261203e-05,
+      "loss": 1.4907,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2739744186401367,
+      "learning_rate": 4.999989426934567e-05,
+      "loss": 1.3993,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7271506786346436,
+      "learning_rate": 4.999989162607932e-05,
+      "loss": 1.6001,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.103022336959839,
+      "learning_rate": 4.9999888982812955e-05,
+      "loss": 1.4349,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.593957901000977,
+      "learning_rate": 4.99998863395466e-05,
+      "loss": 1.9332,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0963082313537598,
+      "learning_rate": 4.999988369628024e-05,
+      "loss": 1.4228,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.47084903717041,
+      "learning_rate": 4.999988105301388e-05,
+      "loss": 1.7036,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.24398136138916,
+      "learning_rate": 4.999987840974753e-05,
+      "loss": 1.5294,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.529242753982544,
+      "learning_rate": 4.9999875766481165e-05,
+      "loss": 1.5671,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3874130249023438,
+      "learning_rate": 4.999987312321481e-05,
+      "loss": 1.3742,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1134917736053467,
+      "learning_rate": 4.9999870479948445e-05,
+      "loss": 1.4917,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8506301641464233,
+      "learning_rate": 4.999986783668209e-05,
+      "loss": 1.4208,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.382322072982788,
+      "learning_rate": 4.999986519341573e-05,
+      "loss": 1.5821,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.377351999282837,
+      "learning_rate": 4.9999862550149375e-05,
+      "loss": 1.6745,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0077297687530518,
+      "learning_rate": 4.999985990688302e-05,
+      "loss": 1.4879,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9965648651123047,
+      "learning_rate": 4.9999857263616655e-05,
+      "loss": 1.4248,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1459808349609375,
+      "learning_rate": 4.99998546203503e-05,
+      "loss": 2.224,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5657927989959717,
+      "learning_rate": 4.999985197708394e-05,
+      "loss": 1.5719,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5736804008483887,
+      "learning_rate": 4.9999849333817585e-05,
+      "loss": 1.4618,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7099759578704834,
+      "learning_rate": 4.999984669055122e-05,
+      "loss": 1.5586,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.536343574523926,
+      "learning_rate": 4.9999844047284864e-05,
+      "loss": 1.655,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.007676839828491,
+      "learning_rate": 4.999984140401851e-05,
+      "loss": 1.4985,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2398571968078613,
+      "learning_rate": 4.999983876075215e-05,
+      "loss": 1.6183,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4275360107421875,
+      "learning_rate": 4.9999836117485794e-05,
+      "loss": 1.5383,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3246569633483887,
+      "learning_rate": 4.999983347421943e-05,
+      "loss": 1.6069,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3558106422424316,
+      "learning_rate": 4.9999830830953074e-05,
+      "loss": 1.5856,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7785794734954834,
+      "learning_rate": 4.999982818768672e-05,
+      "loss": 1.4839,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.317659616470337,
+      "learning_rate": 4.999982554442036e-05,
+      "loss": 1.6549,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4520413875579834,
+      "learning_rate": 4.9999822901154e-05,
+      "loss": 1.5752,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0758883953094482,
+      "learning_rate": 4.999982025788764e-05,
+      "loss": 1.4306,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7527599334716797,
+      "learning_rate": 4.999981761462128e-05,
+      "loss": 1.455,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.283609628677368,
+      "learning_rate": 4.999981497135493e-05,
+      "loss": 1.5723,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8821215629577637,
+      "learning_rate": 4.999981232808857e-05,
+      "loss": 1.7977,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.368835926055908,
+      "learning_rate": 4.999980968482221e-05,
+      "loss": 1.4298,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9589816331863403,
+      "learning_rate": 4.999980704155585e-05,
+      "loss": 1.4099,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.269073486328125,
+      "learning_rate": 4.999980439828949e-05,
+      "loss": 1.5414,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0993194580078125,
+      "learning_rate": 4.999980175502314e-05,
+      "loss": 1.5344,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8945457935333252,
+      "learning_rate": 4.9999799111756774e-05,
+      "loss": 1.5647,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2035017013549805,
+      "learning_rate": 4.999979646849042e-05,
+      "loss": 1.5028,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4101955890655518,
+      "learning_rate": 4.999979382522406e-05,
+      "loss": 1.6601,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.757524013519287,
+      "learning_rate": 4.99997911819577e-05,
+      "loss": 1.4718,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9790518283843994,
+      "learning_rate": 4.999978853869135e-05,
+      "loss": 1.3448,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.137676954269409,
+      "learning_rate": 4.9999785895424984e-05,
+      "loss": 1.6444,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.180936098098755,
+      "learning_rate": 4.999978325215863e-05,
+      "loss": 1.5443,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8605648279190063,
+      "learning_rate": 4.9999780608892264e-05,
+      "loss": 1.4829,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4552361965179443,
+      "learning_rate": 4.999977796562591e-05,
+      "loss": 1.6868,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9902219772338867,
+      "learning_rate": 4.999977532235955e-05,
+      "loss": 1.497,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4461376667022705,
+      "learning_rate": 4.9999772679093194e-05,
+      "loss": 1.4863,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.226517915725708,
+      "learning_rate": 4.999977003582684e-05,
+      "loss": 1.4684,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.703000545501709,
+      "learning_rate": 4.9999767392560474e-05,
+      "loss": 2.6195,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.329559564590454,
+      "learning_rate": 4.999976474929412e-05,
+      "loss": 1.5049,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5275020599365234,
+      "learning_rate": 4.999976210602776e-05,
+      "loss": 1.4523,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.621319532394409,
+      "learning_rate": 4.9999759462761404e-05,
+      "loss": 1.4475,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8720791339874268,
+      "learning_rate": 4.999975681949504e-05,
+      "loss": 1.7144,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0989532470703125,
+      "learning_rate": 4.9999754176228684e-05,
+      "loss": 1.7115,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8010573387145996,
+      "learning_rate": 4.999975153296233e-05,
+      "loss": 1.5537,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8930305242538452,
+      "learning_rate": 4.999974888969597e-05,
+      "loss": 1.4551,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.055920362472534,
+      "learning_rate": 4.9999746246429614e-05,
+      "loss": 1.5935,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8449063301086426,
+      "learning_rate": 4.999974360316325e-05,
+      "loss": 1.3382,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5688421726226807,
+      "learning_rate": 4.9999740959896894e-05,
+      "loss": 1.5015,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0129003524780273,
+      "learning_rate": 4.999973831663054e-05,
+      "loss": 1.4116,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.893972396850586,
+      "learning_rate": 4.999973567336418e-05,
+      "loss": 1.3856,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.010422706604004,
+      "learning_rate": 4.999973303009782e-05,
+      "loss": 1.5988,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6558338403701782,
+      "learning_rate": 4.999973038683146e-05,
+      "loss": 1.4088,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9286550283432007,
+      "learning_rate": 4.9999727743565104e-05,
+      "loss": 1.5706,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8035117387771606,
+      "learning_rate": 4.999972510029875e-05,
+      "loss": 1.5175,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.912644624710083,
+      "learning_rate": 4.999972245703239e-05,
+      "loss": 1.4338,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2979304790496826,
+      "learning_rate": 4.999971981376603e-05,
+      "loss": 1.3808,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.205592632293701,
+      "learning_rate": 4.999971717049967e-05,
+      "loss": 1.7175,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2227728366851807,
+      "learning_rate": 4.9999714527233313e-05,
+      "loss": 1.7261,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2743029594421387,
+      "learning_rate": 4.999971188396696e-05,
+      "loss": 1.793,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9732506275177002,
+      "learning_rate": 4.999970924070059e-05,
+      "loss": 1.6742,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6238417625427246,
+      "learning_rate": 4.999970659743424e-05,
+      "loss": 1.5959,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.796953558921814,
+      "learning_rate": 4.999970395416787e-05,
+      "loss": 1.4698,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3158442974090576,
+      "learning_rate": 4.999970131090152e-05,
+      "loss": 1.565,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.876071572303772,
+      "learning_rate": 4.999969866763517e-05,
+      "loss": 1.4447,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0669713020324707,
+      "learning_rate": 4.99996960243688e-05,
+      "loss": 1.5333,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.298532724380493,
+      "learning_rate": 4.9999693381102447e-05,
+      "loss": 1.4935,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0423359870910645,
+      "learning_rate": 4.999969073783608e-05,
+      "loss": 1.891,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5884052515029907,
+      "learning_rate": 4.999968809456973e-05,
+      "loss": 1.5175,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8445727825164795,
+      "learning_rate": 4.999968545130337e-05,
+      "loss": 1.7275,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0354347229003906,
+      "learning_rate": 4.999968280803701e-05,
+      "loss": 1.4503,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5277605056762695,
+      "learning_rate": 4.999968016477065e-05,
+      "loss": 1.4681,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.2009758949279785,
+      "learning_rate": 4.999967752150429e-05,
+      "loss": 1.6093,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2068192958831787,
+      "learning_rate": 4.999967487823794e-05,
+      "loss": 1.4448,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9573097229003906,
+      "learning_rate": 4.999967223497158e-05,
+      "loss": 1.4026,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3770556449890137,
+      "learning_rate": 4.999966959170522e-05,
+      "loss": 1.5669,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6946535110473633,
+      "learning_rate": 4.999966694843886e-05,
+      "loss": 1.3897,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1800200939178467,
+      "learning_rate": 4.99996643051725e-05,
+      "loss": 1.4924,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9927818775177002,
+      "learning_rate": 4.9999661661906146e-05,
+      "loss": 1.4967,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.767853021621704,
+      "learning_rate": 4.999965901863979e-05,
+      "loss": 1.6371,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.450547456741333,
+      "learning_rate": 4.999965637537343e-05,
+      "loss": 1.7525,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.358370542526245,
+      "learning_rate": 4.999965373210707e-05,
+      "loss": 1.5007,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.935003399848938,
+      "learning_rate": 4.999965108884071e-05,
+      "loss": 1.4791,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.265611171722412,
+      "learning_rate": 4.9999648445574356e-05,
+      "loss": 1.6306,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3422000408172607,
+      "learning_rate": 4.9999645802308e-05,
+      "loss": 1.4634,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4289567470550537,
+      "learning_rate": 4.9999643159041636e-05,
+      "loss": 1.5544,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2966523170471191,
+      "learning_rate": 4.999964051577528e-05,
+      "loss": 2.09,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6301220655441284,
+      "learning_rate": 4.999963787250892e-05,
+      "loss": 1.4887,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.507408857345581,
+      "learning_rate": 4.9999635229242566e-05,
+      "loss": 1.5998,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0291426181793213,
+      "learning_rate": 4.999963258597621e-05,
+      "loss": 1.4295,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0515525341033936,
+      "learning_rate": 4.9999629942709846e-05,
+      "loss": 1.6071,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.540252208709717,
+      "learning_rate": 4.999962729944349e-05,
+      "loss": 1.6251,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.064134359359741,
+      "learning_rate": 4.999962465617713e-05,
+      "loss": 1.5705,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.146686553955078,
+      "learning_rate": 4.9999622012910776e-05,
+      "loss": 1.5845,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.847574234008789,
+      "learning_rate": 4.999961936964441e-05,
+      "loss": 1.5958,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9405486583709717,
+      "learning_rate": 4.9999616726378056e-05,
+      "loss": 1.5463,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2617099285125732,
+      "learning_rate": 4.999961408311169e-05,
+      "loss": 1.8593,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0863020420074463,
+      "learning_rate": 4.999961143984534e-05,
+      "loss": 1.5185,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.108503818511963,
+      "learning_rate": 4.9999608796578986e-05,
+      "loss": 1.438,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7381129264831543,
+      "learning_rate": 4.999960615331262e-05,
+      "loss": 2.0779,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.106574773788452,
+      "learning_rate": 4.9999603510046266e-05,
+      "loss": 1.381,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1027116775512695,
+      "learning_rate": 4.99996008667799e-05,
+      "loss": 1.6812,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.94346022605896,
+      "learning_rate": 4.999959822351355e-05,
+      "loss": 1.4129,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1212854385375977,
+      "learning_rate": 4.999959558024719e-05,
+      "loss": 1.3796,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5647661685943604,
+      "learning_rate": 4.999959293698083e-05,
+      "loss": 1.3157,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8433693647384644,
+      "learning_rate": 4.999959029371447e-05,
+      "loss": 1.4012,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.450104236602783,
+      "learning_rate": 4.999958765044811e-05,
+      "loss": 1.7848,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9799463748931885,
+      "learning_rate": 4.999958500718176e-05,
+      "loss": 1.6082,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8526062965393066,
+      "learning_rate": 4.99995823639154e-05,
+      "loss": 1.4145,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0426783561706543,
+      "learning_rate": 4.999957972064904e-05,
+      "loss": 1.5729,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4283318519592285,
+      "learning_rate": 4.999957707738268e-05,
+      "loss": 1.6217,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9258979558944702,
+      "learning_rate": 4.999957443411632e-05,
+      "loss": 1.4869,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3271448612213135,
+      "learning_rate": 4.9999571790849966e-05,
+      "loss": 1.6268,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0487473011016846,
+      "learning_rate": 4.999956914758361e-05,
+      "loss": 1.4269,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9487249851226807,
+      "learning_rate": 4.9999566504317245e-05,
+      "loss": 1.4507,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.153182029724121,
+      "learning_rate": 4.999956386105089e-05,
+      "loss": 1.7096,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3093857765197754,
+      "learning_rate": 4.999956121778453e-05,
+      "loss": 1.5613,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.981641411781311,
+      "learning_rate": 4.9999558574518175e-05,
+      "loss": 1.4355,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.089085578918457,
+      "learning_rate": 4.999955593125182e-05,
+      "loss": 1.429,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5991677045822144,
+      "learning_rate": 4.9999553287985455e-05,
+      "loss": 1.4492,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.734675884246826,
+      "learning_rate": 4.99995506447191e-05,
+      "loss": 1.5809,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.420393705368042,
+      "learning_rate": 4.999954800145274e-05,
+      "loss": 1.4857,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.31997013092041,
+      "learning_rate": 4.9999545358186385e-05,
+      "loss": 1.5962,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5309725999832153,
+      "learning_rate": 4.999954271492002e-05,
+      "loss": 1.4092,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.847083330154419,
+      "learning_rate": 4.9999540071653665e-05,
+      "loss": 1.6254,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7818189859390259,
+      "learning_rate": 4.999953742838731e-05,
+      "loss": 1.7391,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8894939422607422,
+      "learning_rate": 4.999953478512095e-05,
+      "loss": 1.407,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.986675500869751,
+      "learning_rate": 4.9999532141854595e-05,
+      "loss": 1.5304,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8774054050445557,
+      "learning_rate": 4.999952949858823e-05,
+      "loss": 1.6561,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6687562465667725,
+      "learning_rate": 4.9999526855321875e-05,
+      "loss": 1.4957,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.294642448425293,
+      "learning_rate": 4.999952421205552e-05,
+      "loss": 1.6488,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7567492723464966,
+      "learning_rate": 4.999952156878916e-05,
+      "loss": 1.3182,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7455366849899292,
+      "learning_rate": 4.9999518925522805e-05,
+      "loss": 1.4502,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9700242280960083,
+      "learning_rate": 4.999951628225644e-05,
+      "loss": 1.6122,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.920436382293701,
+      "learning_rate": 4.9999513638990085e-05,
+      "loss": 1.4297,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8701221942901611,
+      "learning_rate": 4.999951099572373e-05,
+      "loss": 1.5964,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0655953884124756,
+      "learning_rate": 4.999950835245737e-05,
+      "loss": 1.4172,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6868557929992676,
+      "learning_rate": 4.999950570919101e-05,
+      "loss": 1.3961,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6082086563110352,
+      "learning_rate": 4.999950306592465e-05,
+      "loss": 1.4216,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9317607879638672,
+      "learning_rate": 4.999950042265829e-05,
+      "loss": 1.3932,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.717310905456543,
+      "learning_rate": 4.999949777939194e-05,
+      "loss": 1.5223,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.811795711517334,
+      "learning_rate": 4.999949513612558e-05,
+      "loss": 1.4126,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.239238739013672,
+      "learning_rate": 4.999949249285922e-05,
+      "loss": 1.5083,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.470919132232666,
+      "learning_rate": 4.999948984959286e-05,
+      "loss": 1.451,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.700941562652588,
+      "learning_rate": 4.99994872063265e-05,
+      "loss": 1.3988,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0511677265167236,
+      "learning_rate": 4.999948456306015e-05,
+      "loss": 1.5319,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6488723754882812,
+      "learning_rate": 4.9999481919793785e-05,
+      "loss": 1.6021,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5556291341781616,
+      "learning_rate": 4.999947927652743e-05,
+      "loss": 1.2611,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8086776733398438,
+      "learning_rate": 4.9999476633261065e-05,
+      "loss": 1.5305,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9255362749099731,
+      "learning_rate": 4.999947398999471e-05,
+      "loss": 2.7744,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9253562688827515,
+      "learning_rate": 4.999947134672836e-05,
+      "loss": 1.2599,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7918652296066284,
+      "learning_rate": 4.9999468703461995e-05,
+      "loss": 1.4864,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7830438613891602,
+      "learning_rate": 4.999946606019564e-05,
+      "loss": 1.5667,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4832141399383545,
+      "learning_rate": 4.9999463416929275e-05,
+      "loss": 1.4987,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9382109642028809,
+      "learning_rate": 4.999946077366292e-05,
+      "loss": 1.6334,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1438534259796143,
+      "learning_rate": 4.999945813039656e-05,
+      "loss": 1.5843,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3544082641601562,
+      "learning_rate": 4.9999455487130205e-05,
+      "loss": 1.4414,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5326969623565674,
+      "learning_rate": 4.999945284386384e-05,
+      "loss": 2.7421,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.546694040298462,
+      "learning_rate": 4.9999450200597485e-05,
+      "loss": 1.4416,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7090721130371094,
+      "learning_rate": 4.999944755733113e-05,
+      "loss": 1.274,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9729373455047607,
+      "learning_rate": 4.999944491406477e-05,
+      "loss": 1.6772,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7716834545135498,
+      "learning_rate": 4.9999442270798415e-05,
+      "loss": 2.3296,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6714955568313599,
+      "learning_rate": 4.999943962753205e-05,
+      "loss": 1.6719,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.038991928100586,
+      "learning_rate": 4.9999436984265694e-05,
+      "loss": 1.5408,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0426888465881348,
+      "learning_rate": 4.999943434099934e-05,
+      "loss": 1.5963,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.477409601211548,
+      "learning_rate": 4.999943169773298e-05,
+      "loss": 1.4044,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.064354658126831,
+      "learning_rate": 4.999942905446662e-05,
+      "loss": 1.4538,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0638487339019775,
+      "learning_rate": 4.999942641120026e-05,
+      "loss": 1.617,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.679330587387085,
+      "learning_rate": 4.9999423767933904e-05,
+      "loss": 1.3266,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.92036771774292,
+      "learning_rate": 4.999942112466755e-05,
+      "loss": 1.514,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8969709873199463,
+      "learning_rate": 4.999941848140119e-05,
+      "loss": 1.5173,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.212925434112549,
+      "learning_rate": 4.999941583813483e-05,
+      "loss": 1.527,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9756252765655518,
+      "learning_rate": 4.999941319486847e-05,
+      "loss": 1.5038,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0038034915924072,
+      "learning_rate": 4.999941055160211e-05,
+      "loss": 1.4745,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9578697681427002,
+      "learning_rate": 4.999940790833576e-05,
+      "loss": 1.5021,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6040619611740112,
+      "learning_rate": 4.9999405265069394e-05,
+      "loss": 1.4716,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.293952465057373,
+      "learning_rate": 4.999940262180304e-05,
+      "loss": 1.5015,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0029995441436768,
+      "learning_rate": 4.999939997853668e-05,
+      "loss": 1.5904,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.083226203918457,
+      "learning_rate": 4.999939733527032e-05,
+      "loss": 1.5463,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.230170965194702,
+      "learning_rate": 4.999939469200397e-05,
+      "loss": 1.6439,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7591358423233032,
+      "learning_rate": 4.9999392048737604e-05,
+      "loss": 1.4599,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8156698942184448,
+      "learning_rate": 4.999938940547125e-05,
+      "loss": 1.5577,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4956191778182983,
+      "learning_rate": 4.9999386762204884e-05,
+      "loss": 1.4419,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9771087169647217,
+      "learning_rate": 4.999938411893853e-05,
+      "loss": 1.5085,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.274364948272705,
+      "learning_rate": 4.999938147567218e-05,
+      "loss": 1.7869,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5949944257736206,
+      "learning_rate": 4.9999378832405814e-05,
+      "loss": 1.365,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1960229873657227,
+      "learning_rate": 4.999937618913946e-05,
+      "loss": 1.5133,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7716684341430664,
+      "learning_rate": 4.9999373545873094e-05,
+      "loss": 1.4592,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.734731674194336,
+      "learning_rate": 4.999937090260674e-05,
+      "loss": 1.3971,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8094791173934937,
+      "learning_rate": 4.999936825934038e-05,
+      "loss": 1.5852,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9565927982330322,
+      "learning_rate": 4.9999365616074024e-05,
+      "loss": 1.5844,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5293364524841309,
+      "learning_rate": 4.999936297280766e-05,
+      "loss": 1.4522,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7323155403137207,
+      "learning_rate": 4.9999360329541304e-05,
+      "loss": 1.3537,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0965566635131836,
+      "learning_rate": 4.999935768627495e-05,
+      "loss": 1.5371,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5242040157318115,
+      "learning_rate": 4.999935504300859e-05,
+      "loss": 1.3985,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1012749671936035,
+      "learning_rate": 4.9999352399742234e-05,
+      "loss": 1.3778,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2793333530426025,
+      "learning_rate": 4.999934975647587e-05,
+      "loss": 2.0804,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8468409776687622,
+      "learning_rate": 4.9999347113209514e-05,
+      "loss": 1.5211,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2302300930023193,
+      "learning_rate": 4.999934446994316e-05,
+      "loss": 1.5591,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.708184003829956,
+      "learning_rate": 4.99993418266768e-05,
+      "loss": 1.3532,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8321616649627686,
+      "learning_rate": 4.999933918341044e-05,
+      "loss": 1.4032,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.457170248031616,
+      "learning_rate": 4.999933654014408e-05,
+      "loss": 1.6535,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0911595821380615,
+      "learning_rate": 4.9999333896877724e-05,
+      "loss": 1.6297,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6577119827270508,
+      "learning_rate": 4.999933125361137e-05,
+      "loss": 1.4231,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.620975136756897,
+      "learning_rate": 4.999932861034501e-05,
+      "loss": 1.4637,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.977067232131958,
+      "learning_rate": 4.999932596707865e-05,
+      "loss": 1.4728,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0209245681762695,
+      "learning_rate": 4.999932332381229e-05,
+      "loss": 1.4832,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.697238564491272,
+      "learning_rate": 4.9999320680545934e-05,
+      "loss": 1.4891,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.014462471008301,
+      "learning_rate": 4.999931803727958e-05,
+      "loss": 1.5421,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.056257963180542,
+      "learning_rate": 4.9999315394013213e-05,
+      "loss": 1.5913,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.341259479522705,
+      "learning_rate": 4.999931275074686e-05,
+      "loss": 1.473,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8077305555343628,
+      "learning_rate": 4.99993101074805e-05,
+      "loss": 2.268,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0249528884887695,
+      "learning_rate": 4.9999307464214143e-05,
+      "loss": 1.3545,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.364046335220337,
+      "learning_rate": 4.999930482094779e-05,
+      "loss": 1.5331,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6941511631011963,
+      "learning_rate": 4.999930217768142e-05,
+      "loss": 1.4845,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5793673992156982,
+      "learning_rate": 4.999929953441507e-05,
+      "loss": 1.3029,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.953776240348816,
+      "learning_rate": 4.99992968911487e-05,
+      "loss": 1.5851,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9646327495574951,
+      "learning_rate": 4.999929424788235e-05,
+      "loss": 1.5401,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9371438026428223,
+      "learning_rate": 4.999929160461599e-05,
+      "loss": 1.4654,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6834771633148193,
+      "learning_rate": 4.999928896134963e-05,
+      "loss": 1.3223,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7947484254837036,
+      "learning_rate": 4.999928631808328e-05,
+      "loss": 1.4202,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7446882724761963,
+      "learning_rate": 4.999928367481691e-05,
+      "loss": 2.722,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5716474056243896,
+      "learning_rate": 4.999928103155056e-05,
+      "loss": 1.5353,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.014468193054199,
+      "learning_rate": 4.99992783882842e-05,
+      "loss": 1.5229,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4449923038482666,
+      "learning_rate": 4.999927574501784e-05,
+      "loss": 1.318,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8352084159851074,
+      "learning_rate": 4.999927310175148e-05,
+      "loss": 1.4962,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7790042161941528,
+      "learning_rate": 4.999927045848512e-05,
+      "loss": 1.5356,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.363827705383301,
+      "learning_rate": 4.9999267815218766e-05,
+      "loss": 1.6112,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5928308963775635,
+      "learning_rate": 4.999926517195241e-05,
+      "loss": 1.7959,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6353635787963867,
+      "learning_rate": 4.999926252868605e-05,
+      "loss": 1.446,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.458118200302124,
+      "learning_rate": 4.999925988541969e-05,
+      "loss": 1.3969,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.629974365234375,
+      "learning_rate": 4.999925724215333e-05,
+      "loss": 1.3904,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2938826084136963,
+      "learning_rate": 4.9999254598886976e-05,
+      "loss": 1.5822,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7931361198425293,
+      "learning_rate": 4.999925195562062e-05,
+      "loss": 1.4591,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.030017375946045,
+      "learning_rate": 4.9999249312354256e-05,
+      "loss": 1.6292,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7302919626235962,
+      "learning_rate": 4.99992466690879e-05,
+      "loss": 1.5254,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.349557876586914,
+      "learning_rate": 4.999924402582154e-05,
+      "loss": 2.3797,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4057705402374268,
+      "learning_rate": 4.9999241382555186e-05,
+      "loss": 1.7262,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2688748836517334,
+      "learning_rate": 4.999923873928883e-05,
+      "loss": 1.4206,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2801005840301514,
+      "learning_rate": 4.9999236096022466e-05,
+      "loss": 1.5822,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.81279718875885,
+      "learning_rate": 4.999923345275611e-05,
+      "loss": 1.3315,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.2440056800842285,
+      "learning_rate": 4.999923080948975e-05,
+      "loss": 1.8072,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.293469190597534,
+      "learning_rate": 4.9999228166223396e-05,
+      "loss": 1.5538,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9121640920639038,
+      "learning_rate": 4.999922552295703e-05,
+      "loss": 1.5556,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.25481915473938,
+      "learning_rate": 4.9999222879690676e-05,
+      "loss": 1.4961,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.962022066116333,
+      "learning_rate": 4.999922023642432e-05,
+      "loss": 1.6076,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9455184936523438,
+      "learning_rate": 4.999921759315796e-05,
+      "loss": 1.6768,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8146153688430786,
+      "learning_rate": 4.9999214949891606e-05,
+      "loss": 1.5735,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.943361520767212,
+      "learning_rate": 4.999921230662524e-05,
+      "loss": 1.484,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4305094480514526,
+      "learning_rate": 4.9999209663358886e-05,
+      "loss": 1.5315,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6350889205932617,
+      "learning_rate": 4.999920702009252e-05,
+      "loss": 1.4305,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6531808376312256,
+      "learning_rate": 4.999920437682617e-05,
+      "loss": 1.3641,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7430516481399536,
+      "learning_rate": 4.999920173355981e-05,
+      "loss": 1.4648,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.83639395236969,
+      "learning_rate": 4.999919909029345e-05,
+      "loss": 1.4048,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8728502988815308,
+      "learning_rate": 4.9999196447027096e-05,
+      "loss": 1.4558,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.659745693206787,
+      "learning_rate": 4.999919380376073e-05,
+      "loss": 1.4885,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.626731276512146,
+      "learning_rate": 4.999919116049438e-05,
+      "loss": 1.4711,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4510924816131592,
+      "learning_rate": 4.999918851722802e-05,
+      "loss": 1.3163,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7017979621887207,
+      "learning_rate": 4.999918587396166e-05,
+      "loss": 1.4731,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.658976435661316,
+      "learning_rate": 4.99991832306953e-05,
+      "loss": 1.4488,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.676328420639038,
+      "learning_rate": 4.999918058742894e-05,
+      "loss": 1.6087,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5237832069396973,
+      "learning_rate": 4.9999177944162586e-05,
+      "loss": 1.5832,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.28145170211792,
+      "learning_rate": 4.999917530089623e-05,
+      "loss": 1.5939,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0533783435821533,
+      "learning_rate": 4.999917265762987e-05,
+      "loss": 1.4882,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3530943393707275,
+      "learning_rate": 4.999917001436351e-05,
+      "loss": 1.6107,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.37731671333313,
+      "learning_rate": 4.999916737109715e-05,
+      "loss": 1.7011,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7806792259216309,
+      "learning_rate": 4.9999164727830796e-05,
+      "loss": 1.3922,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9741883277893066,
+      "learning_rate": 4.999916208456444e-05,
+      "loss": 1.571,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7171446084976196,
+      "learning_rate": 4.9999159441298075e-05,
+      "loss": 1.4778,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.529051661491394,
+      "learning_rate": 4.999915679803172e-05,
+      "loss": 1.1928,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3755501508712769,
+      "learning_rate": 4.999915415476536e-05,
+      "loss": 1.1658,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.98529314994812,
+      "learning_rate": 4.9999151511499006e-05,
+      "loss": 1.5703,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6316972970962524,
+      "learning_rate": 4.999914886823265e-05,
+      "loss": 1.4204,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.032268524169922,
+      "learning_rate": 4.9999146224966285e-05,
+      "loss": 1.5376,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.743217706680298,
+      "learning_rate": 4.999914358169993e-05,
+      "loss": 1.4928,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9639129638671875,
+      "learning_rate": 4.999914093843357e-05,
+      "loss": 1.234,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4965729713439941,
+      "learning_rate": 4.9999138295167215e-05,
+      "loss": 1.3815,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.119900703430176,
+      "learning_rate": 4.999913565190085e-05,
+      "loss": 1.5642,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4156720638275146,
+      "learning_rate": 4.9999133008634495e-05,
+      "loss": 1.4193,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.469802737236023,
+      "learning_rate": 4.999913036536814e-05,
+      "loss": 1.316,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.856350302696228,
+      "learning_rate": 4.999912772210178e-05,
+      "loss": 1.38,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.884356141090393,
+      "learning_rate": 4.9999125078835425e-05,
+      "loss": 1.4313,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7636878490447998,
+      "learning_rate": 4.999912243556906e-05,
+      "loss": 1.5411,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5664491653442383,
+      "learning_rate": 4.9999119792302705e-05,
+      "loss": 1.4443,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1637308597564697,
+      "learning_rate": 4.999911714903635e-05,
+      "loss": 1.5094,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7296667098999023,
+      "learning_rate": 4.999911450576999e-05,
+      "loss": 1.3445,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0039477348327637,
+      "learning_rate": 4.999911186250363e-05,
+      "loss": 1.5222,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0181169509887695,
+      "learning_rate": 4.999910921923727e-05,
+      "loss": 1.5756,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5162782669067383,
+      "learning_rate": 4.9999106575970915e-05,
+      "loss": 1.5147,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.467000722885132,
+      "learning_rate": 4.999910393270456e-05,
+      "loss": 1.7062,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6281334161758423,
+      "learning_rate": 4.99991012894382e-05,
+      "loss": 1.3059,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9651235342025757,
+      "learning_rate": 4.999909864617184e-05,
+      "loss": 1.4333,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6820709705352783,
+      "learning_rate": 4.999909600290548e-05,
+      "loss": 1.4832,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0924854278564453,
+      "learning_rate": 4.999909335963912e-05,
+      "loss": 1.4381,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2179183959960938,
+      "learning_rate": 4.999909071637277e-05,
+      "loss": 1.6294,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.849017143249512,
+      "learning_rate": 4.9999088073106405e-05,
+      "loss": 1.5849,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.331229329109192,
+      "learning_rate": 4.999908542984005e-05,
+      "loss": 1.3541,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5095014572143555,
+      "learning_rate": 4.999908278657369e-05,
+      "loss": 1.3646,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9050902128219604,
+      "learning_rate": 4.999908014330733e-05,
+      "loss": 1.5449,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5517526865005493,
+      "learning_rate": 4.999907750004098e-05,
+      "loss": 1.3952,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.954657793045044,
+      "learning_rate": 4.9999074856774615e-05,
+      "loss": 1.5246,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.254847288131714,
+      "learning_rate": 4.999907221350826e-05,
+      "loss": 1.6489,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4968459606170654,
+      "learning_rate": 4.9999069570241895e-05,
+      "loss": 1.2771,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0875213146209717,
+      "learning_rate": 4.999906692697554e-05,
+      "loss": 1.5187,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6810837984085083,
+      "learning_rate": 4.999906428370918e-05,
+      "loss": 1.4309,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3495044708251953,
+      "learning_rate": 4.9999061640442825e-05,
+      "loss": 1.2959,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0642073154449463,
+      "learning_rate": 4.999905899717647e-05,
+      "loss": 1.6375,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8212666511535645,
+      "learning_rate": 4.9999056353910105e-05,
+      "loss": 1.4776,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8478472232818604,
+      "learning_rate": 4.999905371064375e-05,
+      "loss": 1.4721,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.798508882522583,
+      "learning_rate": 4.999905106737739e-05,
+      "loss": 1.3373,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7405694723129272,
+      "learning_rate": 4.9999048424111035e-05,
+      "loss": 1.4077,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.036994457244873,
+      "learning_rate": 4.999904578084467e-05,
+      "loss": 1.7484,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2079668045043945,
+      "learning_rate": 4.9999043137578315e-05,
+      "loss": 1.6624,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6705102920532227,
+      "learning_rate": 4.999904049431196e-05,
+      "loss": 1.4176,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.317450761795044,
+      "learning_rate": 4.99990378510456e-05,
+      "loss": 1.3711,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.104743242263794,
+      "learning_rate": 4.9999035207779245e-05,
+      "loss": 1.5613,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6332573890686035,
+      "learning_rate": 4.999903256451288e-05,
+      "loss": 1.4173,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4235080480575562,
+      "learning_rate": 4.9999029921246524e-05,
+      "loss": 1.399,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7853851318359375,
+      "learning_rate": 4.999902727798017e-05,
+      "loss": 1.4576,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7125414609909058,
+      "learning_rate": 4.999902463471381e-05,
+      "loss": 1.4428,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9679752588272095,
+      "learning_rate": 4.999902199144745e-05,
+      "loss": 1.6167,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8285239934921265,
+      "learning_rate": 4.999901934818109e-05,
+      "loss": 2.1042,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6840002536773682,
+      "learning_rate": 4.999901670491473e-05,
+      "loss": 1.31,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.96369469165802,
+      "learning_rate": 4.999901406164838e-05,
+      "loss": 1.4879,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.803420066833496,
+      "learning_rate": 4.999901141838202e-05,
+      "loss": 1.4782,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.224663734436035,
+      "learning_rate": 4.999900877511566e-05,
+      "loss": 1.5677,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4943242073059082,
+      "learning_rate": 4.99990061318493e-05,
+      "loss": 1.3381,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1874823570251465,
+      "learning_rate": 4.999900348858294e-05,
+      "loss": 1.593,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8414877653121948,
+      "learning_rate": 4.999900084531659e-05,
+      "loss": 1.4156,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6664515733718872,
+      "learning_rate": 4.9998998202050224e-05,
+      "loss": 1.3643,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5462859869003296,
+      "learning_rate": 4.999899555878387e-05,
+      "loss": 1.7216,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7361072301864624,
+      "learning_rate": 4.9998992915517504e-05,
+      "loss": 1.6903,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0157864093780518,
+      "learning_rate": 4.999899027225115e-05,
+      "loss": 1.2708,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6637675762176514,
+      "learning_rate": 4.99989876289848e-05,
+      "loss": 1.3696,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.163496494293213,
+      "learning_rate": 4.9998984985718434e-05,
+      "loss": 1.3063,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3925740718841553,
+      "learning_rate": 4.999898234245208e-05,
+      "loss": 1.3509,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.636331558227539,
+      "learning_rate": 4.9998979699185714e-05,
+      "loss": 1.3968,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4375636577606201,
+      "learning_rate": 4.999897705591936e-05,
+      "loss": 1.2983,
+      "step": 397
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0446574687957764,
+      "learning_rate": 4.9998974412653e-05,
+      "loss": 1.5996,
+      "step": 398
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5698069334030151,
+      "learning_rate": 4.9998971769386644e-05,
+      "loss": 1.5342,
+      "step": 399
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7382967472076416,
+      "learning_rate": 4.999896912612029e-05,
+      "loss": 1.4611,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8057607412338257,
+      "learning_rate": 4.9998966482853924e-05,
+      "loss": 1.2178,
+      "step": 401
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6612305641174316,
+      "learning_rate": 4.999896383958757e-05,
+      "loss": 1.3584,
+      "step": 402
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8475961685180664,
+      "learning_rate": 4.999896119632121e-05,
+      "loss": 1.4113,
+      "step": 403
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.21657133102417,
+      "learning_rate": 4.9998958553054854e-05,
+      "loss": 1.671,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.575716972351074,
+      "learning_rate": 4.999895590978849e-05,
+      "loss": 1.4883,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.530962347984314,
+      "learning_rate": 4.9998953266522134e-05,
+      "loss": 1.563,
+      "step": 406
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.216825246810913,
+      "learning_rate": 4.999895062325578e-05,
+      "loss": 1.6064,
+      "step": 407
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.327296495437622,
+      "learning_rate": 4.999894797998942e-05,
+      "loss": 1.3071,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.526550054550171,
+      "learning_rate": 4.9998945336723064e-05,
+      "loss": 1.3926,
+      "step": 409
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1582725048065186,
+      "learning_rate": 4.99989426934567e-05,
+      "loss": 1.6964,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8541489839553833,
+      "learning_rate": 4.9998940050190344e-05,
+      "loss": 1.3953,
+      "step": 411
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.30549955368042,
+      "learning_rate": 4.999893740692399e-05,
+      "loss": 1.764,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5392282009124756,
+      "learning_rate": 4.999893476365763e-05,
+      "loss": 1.5064,
+      "step": 413
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7486714124679565,
+      "learning_rate": 4.999893212039127e-05,
+      "loss": 1.439,
+      "step": 414
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.347045660018921,
+      "learning_rate": 4.999892947712491e-05,
+      "loss": 1.4229,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3823535442352295,
+      "learning_rate": 4.9998926833858554e-05,
+      "loss": 1.2936,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7325212955474854,
+      "learning_rate": 4.99989241905922e-05,
+      "loss": 1.619,
+      "step": 417
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7263675928115845,
+      "learning_rate": 4.999892154732584e-05,
+      "loss": 1.5015,
+      "step": 418
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.637427568435669,
+      "learning_rate": 4.999891890405948e-05,
+      "loss": 1.4528,
+      "step": 419
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2784409523010254,
+      "learning_rate": 4.999891626079312e-05,
+      "loss": 1.532,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0185341835021973,
+      "learning_rate": 4.9998913617526764e-05,
+      "loss": 1.3468,
+      "step": 421
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9354544878005981,
+      "learning_rate": 4.999891097426041e-05,
+      "loss": 1.3265,
+      "step": 422
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.26279354095459,
+      "learning_rate": 4.9998908330994043e-05,
+      "loss": 1.6486,
+      "step": 423
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4739882946014404,
+      "learning_rate": 4.999890568772769e-05,
+      "loss": 1.5383,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0663836002349854,
+      "learning_rate": 4.999890304446132e-05,
+      "loss": 1.4844,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3827534914016724,
+      "learning_rate": 4.9998900401194973e-05,
+      "loss": 1.3942,
+      "step": 426
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5585774183273315,
+      "learning_rate": 4.999889775792862e-05,
+      "loss": 1.4758,
+      "step": 427
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4465296268463135,
+      "learning_rate": 4.9998895114662253e-05,
+      "loss": 1.6781,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3868117332458496,
+      "learning_rate": 4.99988924713959e-05,
+      "loss": 1.6332,
+      "step": 429
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8149529695510864,
+      "learning_rate": 4.999888982812953e-05,
+      "loss": 1.5636,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.818013072013855,
+      "learning_rate": 4.9998887184863183e-05,
+      "loss": 1.501,
+      "step": 431
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7827023267745972,
+      "learning_rate": 4.999888454159682e-05,
+      "loss": 1.3999,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8895608186721802,
+      "learning_rate": 4.999888189833046e-05,
+      "loss": 1.3659,
+      "step": 433
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9790616035461426,
+      "learning_rate": 4.99988792550641e-05,
+      "loss": 1.4773,
+      "step": 434
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6706546545028687,
+      "learning_rate": 4.999887661179774e-05,
+      "loss": 1.264,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9165806770324707,
+      "learning_rate": 4.999887396853139e-05,
+      "loss": 1.5382,
+      "step": 436
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8726650476455688,
+      "learning_rate": 4.999887132526503e-05,
+      "loss": 1.5718,
+      "step": 437
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.575060486793518,
+      "learning_rate": 4.999886868199867e-05,
+      "loss": 1.456,
+      "step": 438
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.905265212059021,
+      "learning_rate": 4.999886603873231e-05,
+      "loss": 1.4464,
+      "step": 439
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.281700372695923,
+      "learning_rate": 4.999886339546595e-05,
+      "loss": 1.6499,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9363112449645996,
+      "learning_rate": 4.9998860752199596e-05,
+      "loss": 1.4776,
+      "step": 441
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9906747341156006,
+      "learning_rate": 4.999885810893324e-05,
+      "loss": 1.5494,
+      "step": 442
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.814429759979248,
+      "learning_rate": 4.9998855465666876e-05,
+      "loss": 1.5261,
+      "step": 443
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1458189487457275,
+      "learning_rate": 4.999885282240052e-05,
+      "loss": 1.7087,
+      "step": 444
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2996509075164795,
+      "learning_rate": 4.999885017913416e-05,
+      "loss": 1.6989,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4062042236328125,
+      "learning_rate": 4.9998847535867806e-05,
+      "loss": 1.546,
+      "step": 446
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.706038236618042,
+      "learning_rate": 4.999884489260145e-05,
+      "loss": 1.473,
+      "step": 447
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7012056112289429,
+      "learning_rate": 4.9998842249335086e-05,
+      "loss": 1.3971,
+      "step": 448
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.555124044418335,
+      "learning_rate": 4.999883960606873e-05,
+      "loss": 1.4611,
+      "step": 449
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.628939151763916,
+      "learning_rate": 4.999883696280237e-05,
+      "loss": 1.5001,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1712963581085205,
+      "learning_rate": 4.9998834319536016e-05,
+      "loss": 1.5208,
+      "step": 451
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5055794715881348,
+      "learning_rate": 4.999883167626966e-05,
+      "loss": 1.4898,
+      "step": 452
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.60610830783844,
+      "learning_rate": 4.9998829033003296e-05,
+      "loss": 1.4213,
+      "step": 453
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9946945905685425,
+      "learning_rate": 4.999882638973694e-05,
+      "loss": 1.3816,
+      "step": 454
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4495370388031006,
+      "learning_rate": 4.999882374647058e-05,
+      "loss": 1.4648,
+      "step": 455
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.316918134689331,
+      "learning_rate": 4.9998821103204226e-05,
+      "loss": 1.5333,
+      "step": 456
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.085571765899658,
+      "learning_rate": 4.999881845993786e-05,
+      "loss": 1.5384,
+      "step": 457
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.630589485168457,
+      "learning_rate": 4.9998815816671506e-05,
+      "loss": 1.3728,
+      "step": 458
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6401982307434082,
+      "learning_rate": 4.999881317340514e-05,
+      "loss": 1.3398,
+      "step": 459
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.816530466079712,
+      "learning_rate": 4.999881053013879e-05,
+      "loss": 1.498,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.30181884765625,
+      "learning_rate": 4.9998807886872436e-05,
+      "loss": 1.3795,
+      "step": 461
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4457385540008545,
+      "learning_rate": 4.999880524360607e-05,
+      "loss": 1.637,
+      "step": 462
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7496565580368042,
+      "learning_rate": 4.9998802600339716e-05,
+      "loss": 1.5845,
+      "step": 463
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2107224464416504,
+      "learning_rate": 4.999879995707335e-05,
+      "loss": 1.6189,
+      "step": 464
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8035690784454346,
+      "learning_rate": 4.9998797313807e-05,
+      "loss": 1.4779,
+      "step": 465
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5716707706451416,
+      "learning_rate": 4.999879467054064e-05,
+      "loss": 1.5,
+      "step": 466
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.618260383605957,
+      "learning_rate": 4.999879202727428e-05,
+      "loss": 1.347,
+      "step": 467
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.039462089538574,
+      "learning_rate": 4.999878938400792e-05,
+      "loss": 1.3722,
+      "step": 468
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6648705005645752,
+      "learning_rate": 4.999878674074156e-05,
+      "loss": 1.281,
+      "step": 469
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9147796630859375,
+      "learning_rate": 4.999878409747521e-05,
+      "loss": 1.3958,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.802927017211914,
+      "learning_rate": 4.999878145420885e-05,
+      "loss": 1.4416,
+      "step": 471
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9650660753250122,
+      "learning_rate": 4.999877881094249e-05,
+      "loss": 1.5072,
+      "step": 472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9995487928390503,
+      "learning_rate": 4.999877616767613e-05,
+      "loss": 1.5929,
+      "step": 473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4388364553451538,
+      "learning_rate": 4.999877352440977e-05,
+      "loss": 1.4441,
+      "step": 474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1389412879943848,
+      "learning_rate": 4.9998770881143416e-05,
+      "loss": 1.5964,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5770952701568604,
+      "learning_rate": 4.999876823787706e-05,
+      "loss": 1.5222,
+      "step": 476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3287453651428223,
+      "learning_rate": 4.9998765594610696e-05,
+      "loss": 1.3724,
+      "step": 477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0664279460906982,
+      "learning_rate": 4.999876295134434e-05,
+      "loss": 1.5227,
+      "step": 478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5038436651229858,
+      "learning_rate": 4.999876030807798e-05,
+      "loss": 1.4743,
+      "step": 479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.563154935836792,
+      "learning_rate": 4.9998757664811626e-05,
+      "loss": 1.4061,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8191063404083252,
+      "learning_rate": 4.999875502154527e-05,
+      "loss": 1.488,
+      "step": 481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2203450202941895,
+      "learning_rate": 4.9998752378278906e-05,
+      "loss": 1.5464,
+      "step": 482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9767942428588867,
+      "learning_rate": 4.999874973501255e-05,
+      "loss": 1.3916,
+      "step": 483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9480667114257812,
+      "learning_rate": 4.999874709174619e-05,
+      "loss": 1.7499,
+      "step": 484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7665108442306519,
+      "learning_rate": 4.9998744448479836e-05,
+      "loss": 1.5805,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.390364646911621,
+      "learning_rate": 4.999874180521347e-05,
+      "loss": 1.3716,
+      "step": 486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.062652826309204,
+      "learning_rate": 4.9998739161947115e-05,
+      "loss": 1.5874,
+      "step": 487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8057736158370972,
+      "learning_rate": 4.999873651868076e-05,
+      "loss": 1.3855,
+      "step": 488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.386656403541565,
+      "learning_rate": 4.99987338754144e-05,
+      "loss": 1.2901,
+      "step": 489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7236733436584473,
+      "learning_rate": 4.9998731232148045e-05,
+      "loss": 1.4226,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0356686115264893,
+      "learning_rate": 4.999872858888168e-05,
+      "loss": 1.5656,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6085858345031738,
+      "learning_rate": 4.9998725945615325e-05,
+      "loss": 1.455,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1326470375061035,
+      "learning_rate": 4.999872330234897e-05,
+      "loss": 1.5803,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9373149871826172,
+      "learning_rate": 4.999872065908261e-05,
+      "loss": 1.468,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.76348876953125,
+      "learning_rate": 4.999871801581625e-05,
+      "loss": 1.4414,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8664847612380981,
+      "learning_rate": 4.999871537254989e-05,
+      "loss": 1.5145,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.551046371459961,
+      "learning_rate": 4.9998712729283535e-05,
+      "loss": 1.4781,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4362707138061523,
+      "learning_rate": 4.999871008601718e-05,
+      "loss": 1.7269,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.114346504211426,
+      "learning_rate": 4.999870744275082e-05,
+      "loss": 1.5489,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2777223587036133,
+      "learning_rate": 4.999870479948446e-05,
+      "loss": 1.4238,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.767372727394104,
+      "learning_rate": 4.99987021562181e-05,
+      "loss": 1.5474,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8725348711013794,
+      "learning_rate": 4.999869951295174e-05,
+      "loss": 1.367,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5889549255371094,
+      "learning_rate": 4.999869686968539e-05,
+      "loss": 1.4077,
+      "step": 503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7493702173233032,
+      "learning_rate": 4.999869422641903e-05,
+      "loss": 1.3959,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4335589408874512,
+      "learning_rate": 4.999869158315267e-05,
+      "loss": 1.4746,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.165184736251831,
+      "learning_rate": 4.999868893988631e-05,
+      "loss": 2.1498,
+      "step": 506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1639416217803955,
+      "learning_rate": 4.999868629661995e-05,
+      "loss": 1.6987,
+      "step": 507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8927098512649536,
+      "learning_rate": 4.99986836533536e-05,
+      "loss": 1.398,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5680326223373413,
+      "learning_rate": 4.9998681010087235e-05,
+      "loss": 1.3932,
+      "step": 509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4867005348205566,
+      "learning_rate": 4.999867836682088e-05,
+      "loss": 1.4149,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.701765537261963,
+      "learning_rate": 4.9998675723554515e-05,
+      "loss": 1.5511,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4900785684585571,
+      "learning_rate": 4.999867308028816e-05,
+      "loss": 1.4106,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6058911085128784,
+      "learning_rate": 4.999867043702181e-05,
+      "loss": 1.4829,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6468846797943115,
+      "learning_rate": 4.9998667793755445e-05,
+      "loss": 1.3159,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6792536973953247,
+      "learning_rate": 4.999866515048909e-05,
+      "loss": 1.2909,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9518896341323853,
+      "learning_rate": 4.9998662507222725e-05,
+      "loss": 1.3782,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7944221496582031,
+      "learning_rate": 4.999865986395637e-05,
+      "loss": 1.5385,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5956846475601196,
+      "learning_rate": 4.999865722069001e-05,
+      "loss": 1.6,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.885039210319519,
+      "learning_rate": 4.9998654577423655e-05,
+      "loss": 1.4915,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6242964267730713,
+      "learning_rate": 4.999865193415729e-05,
+      "loss": 1.4279,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3695862293243408,
+      "learning_rate": 4.9998649290890935e-05,
+      "loss": 1.346,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7382376194000244,
+      "learning_rate": 4.999864664762458e-05,
+      "loss": 1.4477,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.093362808227539,
+      "learning_rate": 4.999864400435822e-05,
+      "loss": 1.5935,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2082191705703735,
+      "learning_rate": 4.9998641361091865e-05,
+      "loss": 1.2539,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1027026176452637,
+      "learning_rate": 4.99986387178255e-05,
+      "loss": 1.5584,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9960774183273315,
+      "learning_rate": 4.9998636074559145e-05,
+      "loss": 1.7921,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.611415147781372,
+      "learning_rate": 4.999863343129279e-05,
+      "loss": 1.4429,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4030812978744507,
+      "learning_rate": 4.999863078802643e-05,
+      "loss": 1.3746,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7626007795333862,
+      "learning_rate": 4.999862814476007e-05,
+      "loss": 1.3982,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5364214181900024,
+      "learning_rate": 4.999862550149371e-05,
+      "loss": 1.3443,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.621209740638733,
+      "learning_rate": 4.9998622858227355e-05,
+      "loss": 1.4546,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9564650058746338,
+      "learning_rate": 4.9998620214961e-05,
+      "loss": 1.486,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0014400482177734,
+      "learning_rate": 4.999861757169464e-05,
+      "loss": 1.4936,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.072169780731201,
+      "learning_rate": 4.999861492842828e-05,
+      "loss": 1.4523,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.796176552772522,
+      "learning_rate": 4.999861228516192e-05,
+      "loss": 1.4846,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7444260120391846,
+      "learning_rate": 4.999860964189556e-05,
+      "loss": 1.7205,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0753159523010254,
+      "learning_rate": 4.999860699862921e-05,
+      "loss": 1.5154,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.922376036643982,
+      "learning_rate": 4.9998604355362844e-05,
+      "loss": 1.4743,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6779506206512451,
+      "learning_rate": 4.999860171209649e-05,
+      "loss": 1.8775,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8525536060333252,
+      "learning_rate": 4.999859906883013e-05,
+      "loss": 1.503,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.569024920463562,
+      "learning_rate": 4.999859642556377e-05,
+      "loss": 1.492,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.278597116470337,
+      "learning_rate": 4.999859378229742e-05,
+      "loss": 1.6015,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.669932246208191,
+      "learning_rate": 4.9998591139031054e-05,
+      "loss": 1.5715,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1131021976470947,
+      "learning_rate": 4.99985884957647e-05,
+      "loss": 1.4514,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4460028409957886,
+      "learning_rate": 4.9998585852498334e-05,
+      "loss": 1.3824,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8029924631118774,
+      "learning_rate": 4.999858320923198e-05,
+      "loss": 1.4675,
+      "step": 546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.855341911315918,
+      "learning_rate": 4.999858056596562e-05,
+      "loss": 1.5242,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5050597190856934,
+      "learning_rate": 4.9998577922699264e-05,
+      "loss": 1.3618,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.354640245437622,
+      "learning_rate": 4.999857527943291e-05,
+      "loss": 1.3888,
+      "step": 549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.641378402709961,
+      "learning_rate": 4.9998572636166544e-05,
+      "loss": 1.302,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.419879674911499,
+      "learning_rate": 4.999856999290019e-05,
+      "loss": 1.421,
+      "step": 551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6059188842773438,
+      "learning_rate": 4.999856734963383e-05,
+      "loss": 1.2002,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5114516019821167,
+      "learning_rate": 4.9998564706367474e-05,
+      "loss": 1.5001,
+      "step": 553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7145854234695435,
+      "learning_rate": 4.999856206310111e-05,
+      "loss": 1.4532,
+      "step": 554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6657171249389648,
+      "learning_rate": 4.9998559419834754e-05,
+      "loss": 1.3517,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0381858348846436,
+      "learning_rate": 4.99985567765684e-05,
+      "loss": 1.383,
+      "step": 556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2837612628936768,
+      "learning_rate": 4.999855413330204e-05,
+      "loss": 1.3409,
+      "step": 557
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5409398078918457,
+      "learning_rate": 4.9998551490035684e-05,
+      "loss": 2.4106,
+      "step": 558
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6652171611785889,
+      "learning_rate": 4.999854884676932e-05,
+      "loss": 1.3878,
+      "step": 559
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.811204433441162,
+      "learning_rate": 4.9998546203502964e-05,
+      "loss": 1.4947,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8710405826568604,
+      "learning_rate": 4.999854356023661e-05,
+      "loss": 1.4688,
+      "step": 561
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.472464680671692,
+      "learning_rate": 4.999854091697025e-05,
+      "loss": 1.3722,
+      "step": 562
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6347790956497192,
+      "learning_rate": 4.999853827370389e-05,
+      "loss": 1.4845,
+      "step": 563
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9538577795028687,
+      "learning_rate": 4.999853563043753e-05,
+      "loss": 1.5164,
+      "step": 564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4320762157440186,
+      "learning_rate": 4.9998532987171174e-05,
+      "loss": 1.3518,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8663872480392456,
+      "learning_rate": 4.999853034390482e-05,
+      "loss": 1.3925,
+      "step": 566
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3366427421569824,
+      "learning_rate": 4.999852770063846e-05,
+      "loss": 1.5293,
+      "step": 567
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9861294031143188,
+      "learning_rate": 4.99985250573721e-05,
+      "loss": 1.603,
+      "step": 568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5574374198913574,
+      "learning_rate": 4.999852241410574e-05,
+      "loss": 1.3909,
+      "step": 569
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3627758026123047,
+      "learning_rate": 4.9998519770839384e-05,
+      "loss": 1.4285,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7316529750823975,
+      "learning_rate": 4.999851712757303e-05,
+      "loss": 1.5214,
+      "step": 571
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7259809970855713,
+      "learning_rate": 4.9998514484306664e-05,
+      "loss": 1.5174,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.413907766342163,
+      "learning_rate": 4.999851184104031e-05,
+      "loss": 1.4121,
+      "step": 573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7596582174301147,
+      "learning_rate": 4.999850919777395e-05,
+      "loss": 1.2815,
+      "step": 574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6428536176681519,
+      "learning_rate": 4.9998506554507594e-05,
+      "loss": 1.5396,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4950534105300903,
+      "learning_rate": 4.999850391124124e-05,
+      "loss": 1.4607,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9184705018997192,
+      "learning_rate": 4.9998501267974874e-05,
+      "loss": 1.562,
+      "step": 577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7720540761947632,
+      "learning_rate": 4.999849862470852e-05,
+      "loss": 1.6404,
+      "step": 578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0816500186920166,
+      "learning_rate": 4.9998495981442153e-05,
+      "loss": 1.4943,
+      "step": 579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9151034355163574,
+      "learning_rate": 4.9998493338175804e-05,
+      "loss": 1.5095,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1822967529296875,
+      "learning_rate": 4.999849069490944e-05,
+      "loss": 1.5597,
+      "step": 581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9017130136489868,
+      "learning_rate": 4.9998488051643083e-05,
+      "loss": 1.3345,
+      "step": 582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4737253189086914,
+      "learning_rate": 4.999848540837673e-05,
+      "loss": 1.4979,
+      "step": 583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2634191513061523,
+      "learning_rate": 4.999848276511036e-05,
+      "loss": 2.8454,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.760047435760498,
+      "learning_rate": 4.9998480121844013e-05,
+      "loss": 1.4509,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6695431470870972,
+      "learning_rate": 4.999847747857765e-05,
+      "loss": 1.4223,
+      "step": 586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3111886978149414,
+      "learning_rate": 4.999847483531129e-05,
+      "loss": 1.5982,
+      "step": 587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8031593561172485,
+      "learning_rate": 4.999847219204493e-05,
+      "loss": 1.4967,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9424952268600464,
+      "learning_rate": 4.999846954877857e-05,
+      "loss": 1.324,
+      "step": 589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2639620304107666,
+      "learning_rate": 4.9998466905512217e-05,
+      "loss": 1.3666,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6516600847244263,
+      "learning_rate": 4.999846426224586e-05,
+      "loss": 1.4514,
+      "step": 591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4865288734436035,
+      "learning_rate": 4.99984616189795e-05,
+      "loss": 1.5495,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5467456579208374,
+      "learning_rate": 4.999845897571314e-05,
+      "loss": 1.486,
+      "step": 593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.348191499710083,
+      "learning_rate": 4.999845633244678e-05,
+      "loss": 1.5968,
+      "step": 594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7877027988433838,
+      "learning_rate": 4.9998453689180426e-05,
+      "loss": 1.3592,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5933834314346313,
+      "learning_rate": 4.999845104591407e-05,
+      "loss": 1.4955,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6282588243484497,
+      "learning_rate": 4.9998448402647706e-05,
+      "loss": 1.4319,
+      "step": 597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.644222617149353,
+      "learning_rate": 4.999844575938135e-05,
+      "loss": 1.3771,
+      "step": 598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6014835834503174,
+      "learning_rate": 4.999844311611499e-05,
+      "loss": 1.3683,
+      "step": 599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8688406944274902,
+      "learning_rate": 4.9998440472848636e-05,
+      "loss": 1.5353,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8344178199768066,
+      "learning_rate": 4.999843782958228e-05,
+      "loss": 1.544,
+      "step": 601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.972238540649414,
+      "learning_rate": 4.9998435186315916e-05,
+      "loss": 1.6611,
+      "step": 602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.057040214538574,
+      "learning_rate": 4.999843254304956e-05,
+      "loss": 1.5153,
+      "step": 603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.679627776145935,
+      "learning_rate": 4.99984298997832e-05,
+      "loss": 1.4062,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3431954383850098,
+      "learning_rate": 4.9998427256516846e-05,
+      "loss": 1.3,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0472002029418945,
+      "learning_rate": 4.999842461325048e-05,
+      "loss": 1.4175,
+      "step": 606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.782632827758789,
+      "learning_rate": 4.9998421969984126e-05,
+      "loss": 1.6073,
+      "step": 607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4096276760101318,
+      "learning_rate": 4.999841932671777e-05,
+      "loss": 1.4098,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7201117277145386,
+      "learning_rate": 4.999841668345141e-05,
+      "loss": 1.5576,
+      "step": 609
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7371008396148682,
+      "learning_rate": 4.9998414040185056e-05,
+      "loss": 1.5681,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.920829176902771,
+      "learning_rate": 4.999841139691869e-05,
+      "loss": 1.4491,
+      "step": 611
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6799613237380981,
+      "learning_rate": 4.9998408753652336e-05,
+      "loss": 1.321,
+      "step": 612
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5682315826416016,
+      "learning_rate": 4.999840611038597e-05,
+      "loss": 1.518,
+      "step": 613
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3978716135025024,
+      "learning_rate": 4.999840346711962e-05,
+      "loss": 1.3173,
+      "step": 614
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.726836919784546,
+      "learning_rate": 4.999840082385326e-05,
+      "loss": 1.2722,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4490840435028076,
+      "learning_rate": 4.99983981805869e-05,
+      "loss": 1.4889,
+      "step": 616
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7568596601486206,
+      "learning_rate": 4.9998395537320546e-05,
+      "loss": 1.4461,
+      "step": 617
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0599470138549805,
+      "learning_rate": 4.999839289405418e-05,
+      "loss": 1.5242,
+      "step": 618
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7950060367584229,
+      "learning_rate": 4.999839025078783e-05,
+      "loss": 1.5163,
+      "step": 619
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4780685901641846,
+      "learning_rate": 4.999838760752147e-05,
+      "loss": 1.3264,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6327966451644897,
+      "learning_rate": 4.999838496425511e-05,
+      "loss": 1.4667,
+      "step": 621
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.074615001678467,
+      "learning_rate": 4.999838232098875e-05,
+      "loss": 1.511,
+      "step": 622
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8744715452194214,
+      "learning_rate": 4.999837967772239e-05,
+      "loss": 1.4252,
+      "step": 623
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.704593300819397,
+      "learning_rate": 4.9998377034456036e-05,
+      "loss": 1.3698,
+      "step": 624
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8473715782165527,
+      "learning_rate": 4.999837439118968e-05,
+      "loss": 1.7187,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2067062854766846,
+      "learning_rate": 4.999837174792332e-05,
+      "loss": 1.538,
+      "step": 626
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7628772258758545,
+      "learning_rate": 4.999836910465696e-05,
+      "loss": 1.8225,
+      "step": 627
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9866862297058105,
+      "learning_rate": 4.99983664613906e-05,
+      "loss": 1.5987,
+      "step": 628
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.985856533050537,
+      "learning_rate": 4.9998363818124246e-05,
+      "loss": 1.7225,
+      "step": 629
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7447030544281006,
+      "learning_rate": 4.999836117485789e-05,
+      "loss": 1.5741,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9388636350631714,
+      "learning_rate": 4.9998358531591526e-05,
+      "loss": 1.5376,
+      "step": 631
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.579190731048584,
+      "learning_rate": 4.999835588832517e-05,
+      "loss": 1.7387,
+      "step": 632
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4859349727630615,
+      "learning_rate": 4.999835324505881e-05,
+      "loss": 1.9736,
+      "step": 633
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1519176959991455,
+      "learning_rate": 4.9998350601792456e-05,
+      "loss": 1.5654,
+      "step": 634
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.412064552307129,
+      "learning_rate": 4.99983479585261e-05,
+      "loss": 1.5882,
+      "step": 635
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5947344303131104,
+      "learning_rate": 4.9998345315259736e-05,
+      "loss": 1.8252,
+      "step": 636
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6512680053710938,
+      "learning_rate": 4.999834267199338e-05,
+      "loss": 1.5287,
+      "step": 637
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.122990369796753,
+      "learning_rate": 4.999834002872702e-05,
+      "loss": 1.6539,
+      "step": 638
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5424320697784424,
+      "learning_rate": 4.9998337385460666e-05,
+      "loss": 1.4815,
+      "step": 639
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5155887603759766,
+      "learning_rate": 4.99983347421943e-05,
+      "loss": 1.5148,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.436986207962036,
+      "learning_rate": 4.9998332098927945e-05,
+      "loss": 1.5537,
+      "step": 641
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7735384702682495,
+      "learning_rate": 4.999832945566159e-05,
+      "loss": 1.3864,
+      "step": 642
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.07029128074646,
+      "learning_rate": 4.999832681239523e-05,
+      "loss": 1.5162,
+      "step": 643
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2494288682937622,
+      "learning_rate": 4.9998324169128875e-05,
+      "loss": 1.4189,
+      "step": 644
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8730602264404297,
+      "learning_rate": 4.999832152586251e-05,
+      "loss": 1.4463,
+      "step": 645
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.662687063217163,
+      "learning_rate": 4.9998318882596155e-05,
+      "loss": 1.4679,
+      "step": 646
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6732896566390991,
+      "learning_rate": 4.99983162393298e-05,
+      "loss": 1.4902,
+      "step": 647
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.843627691268921,
+      "learning_rate": 4.999831359606344e-05,
+      "loss": 1.5789,
+      "step": 648
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6948591470718384,
+      "learning_rate": 4.999831095279708e-05,
+      "loss": 1.5464,
+      "step": 649
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.428767442703247,
+      "learning_rate": 4.999830830953072e-05,
+      "loss": 1.3805,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4716472625732422,
+      "learning_rate": 4.999830566626436e-05,
+      "loss": 1.3735,
+      "step": 651
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9272854328155518,
+      "learning_rate": 4.999830302299801e-05,
+      "loss": 1.4985,
+      "step": 652
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4778728485107422,
+      "learning_rate": 4.999830037973165e-05,
+      "loss": 1.5222,
+      "step": 653
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.656208872795105,
+      "learning_rate": 4.999829773646529e-05,
+      "loss": 1.5203,
+      "step": 654
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4637503623962402,
+      "learning_rate": 4.999829509319893e-05,
+      "loss": 1.3019,
+      "step": 655
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2546286582946777,
+      "learning_rate": 4.999829244993257e-05,
+      "loss": 1.4445,
+      "step": 656
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8981226682662964,
+      "learning_rate": 4.999828980666622e-05,
+      "loss": 1.5292,
+      "step": 657
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7503554821014404,
+      "learning_rate": 4.9998287163399855e-05,
+      "loss": 1.5355,
+      "step": 658
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9481329917907715,
+      "learning_rate": 4.99982845201335e-05,
+      "loss": 1.5635,
+      "step": 659
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9574804306030273,
+      "learning_rate": 4.999828187686714e-05,
+      "loss": 1.5606,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4097222089767456,
+      "learning_rate": 4.999827923360078e-05,
+      "loss": 1.3936,
+      "step": 661
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.316592216491699,
+      "learning_rate": 4.999827659033443e-05,
+      "loss": 2.6892,
+      "step": 662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6841673851013184,
+      "learning_rate": 4.9998273947068065e-05,
+      "loss": 1.5491,
+      "step": 663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2221481800079346,
+      "learning_rate": 4.999827130380171e-05,
+      "loss": 2.0248,
+      "step": 664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9443615674972534,
+      "learning_rate": 4.9998268660535345e-05,
+      "loss": 1.7734,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4951661825180054,
+      "learning_rate": 4.999826601726899e-05,
+      "loss": 1.3432,
+      "step": 666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9404951333999634,
+      "learning_rate": 4.999826337400263e-05,
+      "loss": 1.5152,
+      "step": 667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.332512855529785,
+      "learning_rate": 4.9998260730736275e-05,
+      "loss": 1.4796,
+      "step": 668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5245940685272217,
+      "learning_rate": 4.999825808746992e-05,
+      "loss": 1.6852,
+      "step": 669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.114135980606079,
+      "learning_rate": 4.9998255444203555e-05,
+      "loss": 1.4631,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7418584823608398,
+      "learning_rate": 4.99982528009372e-05,
+      "loss": 1.5489,
+      "step": 671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3780205249786377,
+      "learning_rate": 4.999825015767084e-05,
+      "loss": 1.3947,
+      "step": 672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7192987203598022,
+      "learning_rate": 4.9998247514404485e-05,
+      "loss": 1.4689,
+      "step": 673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8395838737487793,
+      "learning_rate": 4.999824487113812e-05,
+      "loss": 1.8518,
+      "step": 674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4088737964630127,
+      "learning_rate": 4.9998242227871765e-05,
+      "loss": 1.5454,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8842014074325562,
+      "learning_rate": 4.999823958460541e-05,
+      "loss": 1.3811,
+      "step": 676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.678695559501648,
+      "learning_rate": 4.999823694133905e-05,
+      "loss": 1.4557,
+      "step": 677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.40257728099823,
+      "learning_rate": 4.9998234298072695e-05,
+      "loss": 1.3856,
+      "step": 678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.489301085472107,
+      "learning_rate": 4.999823165480633e-05,
+      "loss": 1.4141,
+      "step": 679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7530710697174072,
+      "learning_rate": 4.9998229011539975e-05,
+      "loss": 1.4127,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5724650621414185,
+      "learning_rate": 4.999822636827362e-05,
+      "loss": 1.24,
+      "step": 681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2000969648361206,
+      "learning_rate": 4.999822372500726e-05,
+      "loss": 1.364,
+      "step": 682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8323371410369873,
+      "learning_rate": 4.99982210817409e-05,
+      "loss": 1.49,
+      "step": 683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7359663248062134,
+      "learning_rate": 4.999821843847454e-05,
+      "loss": 2.1725,
+      "step": 684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7871060371398926,
+      "learning_rate": 4.999821579520818e-05,
+      "loss": 1.4377,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.544847011566162,
+      "learning_rate": 4.999821315194183e-05,
+      "loss": 1.6791,
+      "step": 686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7650666236877441,
+      "learning_rate": 4.999821050867547e-05,
+      "loss": 1.2774,
+      "step": 687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6660563945770264,
+      "learning_rate": 4.999820786540911e-05,
+      "loss": 1.4697,
+      "step": 688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.015302896499634,
+      "learning_rate": 4.999820522214275e-05,
+      "loss": 1.3511,
+      "step": 689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5186625719070435,
+      "learning_rate": 4.999820257887639e-05,
+      "loss": 1.3887,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6178382635116577,
+      "learning_rate": 4.999819993561004e-05,
+      "loss": 1.4698,
+      "step": 691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.112569808959961,
+      "learning_rate": 4.9998197292343674e-05,
+      "loss": 1.3505,
+      "step": 692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9174686670303345,
+      "learning_rate": 4.999819464907732e-05,
+      "loss": 1.3709,
+      "step": 693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0415711402893066,
+      "learning_rate": 4.9998192005810954e-05,
+      "loss": 1.4344,
+      "step": 694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6479341983795166,
+      "learning_rate": 4.99981893625446e-05,
+      "loss": 1.5734,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1740458011627197,
+      "learning_rate": 4.999818671927825e-05,
+      "loss": 1.6263,
+      "step": 696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.785251498222351,
+      "learning_rate": 4.9998184076011884e-05,
+      "loss": 1.5224,
+      "step": 697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0787441730499268,
+      "learning_rate": 4.999818143274553e-05,
+      "loss": 1.4544,
+      "step": 698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4064971208572388,
+      "learning_rate": 4.9998178789479164e-05,
+      "loss": 1.226,
+      "step": 699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8769772052764893,
+      "learning_rate": 4.999817614621281e-05,
+      "loss": 1.49,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.69243323802948,
+      "learning_rate": 4.999817350294645e-05,
+      "loss": 1.4533,
+      "step": 701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.887552499771118,
+      "learning_rate": 4.9998170859680094e-05,
+      "loss": 1.5211,
+      "step": 702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5650453567504883,
+      "learning_rate": 4.999816821641373e-05,
+      "loss": 1.3739,
+      "step": 703
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3201148509979248,
+      "learning_rate": 4.9998165573147374e-05,
+      "loss": 1.3814,
+      "step": 704
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.713008165359497,
+      "learning_rate": 4.999816292988102e-05,
+      "loss": 1.6088,
+      "step": 705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5600684881210327,
+      "learning_rate": 4.999816028661466e-05,
+      "loss": 1.4981,
+      "step": 706
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8583405017852783,
+      "learning_rate": 4.9998157643348304e-05,
+      "loss": 1.5151,
+      "step": 707
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3582550287246704,
+      "learning_rate": 4.999815500008194e-05,
+      "loss": 1.2428,
+      "step": 708
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7297612428665161,
+      "learning_rate": 4.9998152356815584e-05,
+      "loss": 1.4642,
+      "step": 709
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7613017559051514,
+      "learning_rate": 4.999814971354923e-05,
+      "loss": 1.5286,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2815446853637695,
+      "learning_rate": 4.999814707028287e-05,
+      "loss": 1.3764,
+      "step": 711
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5956687927246094,
+      "learning_rate": 4.9998144427016514e-05,
+      "loss": 1.5404,
+      "step": 712
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.64644455909729,
+      "learning_rate": 4.999814178375015e-05,
+      "loss": 1.4319,
+      "step": 713
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7653874158859253,
+      "learning_rate": 4.9998139140483794e-05,
+      "loss": 1.4878,
+      "step": 714
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4656555652618408,
+      "learning_rate": 4.999813649721744e-05,
+      "loss": 1.2735,
+      "step": 715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7053130865097046,
+      "learning_rate": 4.999813385395108e-05,
+      "loss": 1.3343,
+      "step": 716
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.536278247833252,
+      "learning_rate": 4.999813121068472e-05,
+      "loss": 1.3648,
+      "step": 717
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.818792700767517,
+      "learning_rate": 4.999812856741836e-05,
+      "loss": 1.4819,
+      "step": 718
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7577495574951172,
+      "learning_rate": 4.9998125924152004e-05,
+      "loss": 1.4814,
+      "step": 719
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8896210193634033,
+      "learning_rate": 4.999812328088565e-05,
+      "loss": 1.495,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.005990505218506,
+      "learning_rate": 4.999812063761929e-05,
+      "loss": 1.4514,
+      "step": 721
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5241402387619019,
+      "learning_rate": 4.999811799435293e-05,
+      "loss": 1.5343,
+      "step": 722
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5354160070419312,
+      "learning_rate": 4.999811535108657e-05,
+      "loss": 1.4942,
+      "step": 723
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.382417917251587,
+      "learning_rate": 4.9998112707820214e-05,
+      "loss": 1.3767,
+      "step": 724
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4042680263519287,
+      "learning_rate": 4.999811006455386e-05,
+      "loss": 1.5656,
+      "step": 725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.28047513961792,
+      "learning_rate": 4.9998107421287494e-05,
+      "loss": 2.179,
+      "step": 726
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5764708518981934,
+      "learning_rate": 4.999810477802114e-05,
+      "loss": 1.6138,
+      "step": 727
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3149834871292114,
+      "learning_rate": 4.9998102134754774e-05,
+      "loss": 1.4381,
+      "step": 728
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.569943904876709,
+      "learning_rate": 4.9998099491488424e-05,
+      "loss": 1.3762,
+      "step": 729
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1190805435180664,
+      "learning_rate": 4.999809684822207e-05,
+      "loss": 1.4092,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.845038652420044,
+      "learning_rate": 4.9998094204955704e-05,
+      "loss": 1.4469,
+      "step": 731
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5646469593048096,
+      "learning_rate": 4.999809156168935e-05,
+      "loss": 1.4352,
+      "step": 732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5966767072677612,
+      "learning_rate": 4.9998088918422983e-05,
+      "loss": 1.4418,
+      "step": 733
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.536919355392456,
+      "learning_rate": 4.9998086275156634e-05,
+      "loss": 1.821,
+      "step": 734
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5810304880142212,
+      "learning_rate": 4.999808363189027e-05,
+      "loss": 1.3036,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.385150194168091,
+      "learning_rate": 4.9998080988623913e-05,
+      "loss": 1.3625,
+      "step": 736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0804502964019775,
+      "learning_rate": 4.999807834535755e-05,
+      "loss": 1.4638,
+      "step": 737
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3375794887542725,
+      "learning_rate": 4.999807570209119e-05,
+      "loss": 1.8872,
+      "step": 738
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7170017957687378,
+      "learning_rate": 4.9998073058824843e-05,
+      "loss": 1.4212,
+      "step": 739
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0184805393218994,
+      "learning_rate": 4.999807041555848e-05,
+      "loss": 1.5698,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.034041404724121,
+      "learning_rate": 4.999806777229212e-05,
+      "loss": 1.421,
+      "step": 741
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5672121047973633,
+      "learning_rate": 4.999806512902576e-05,
+      "loss": 1.2049,
+      "step": 742
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4102325439453125,
+      "learning_rate": 4.99980624857594e-05,
+      "loss": 1.3471,
+      "step": 743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.961679220199585,
+      "learning_rate": 4.9998059842493047e-05,
+      "loss": 1.6595,
+      "step": 744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7889907360076904,
+      "learning_rate": 4.999805719922669e-05,
+      "loss": 1.3945,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9106556177139282,
+      "learning_rate": 4.9998054555960326e-05,
+      "loss": 1.8885,
+      "step": 746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8228586912155151,
+      "learning_rate": 4.999805191269397e-05,
+      "loss": 1.4767,
+      "step": 747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3569839000701904,
+      "learning_rate": 4.999804926942761e-05,
+      "loss": 1.3065,
+      "step": 748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8571410179138184,
+      "learning_rate": 4.9998046626161257e-05,
+      "loss": 1.3525,
+      "step": 749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8837288618087769,
+      "learning_rate": 4.99980439828949e-05,
+      "loss": 1.4507,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4693742990493774,
+      "learning_rate": 4.9998041339628536e-05,
+      "loss": 1.3506,
+      "step": 751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.089855670928955,
+      "learning_rate": 4.999803869636218e-05,
+      "loss": 1.5604,
+      "step": 752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.408683180809021,
+      "learning_rate": 4.999803605309582e-05,
+      "loss": 1.4014,
+      "step": 753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5929317474365234,
+      "learning_rate": 4.9998033409829466e-05,
+      "loss": 1.4101,
+      "step": 754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5175528526306152,
+      "learning_rate": 4.99980307665631e-05,
+      "loss": 1.5469,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.514690637588501,
+      "learning_rate": 4.9998028123296746e-05,
+      "loss": 1.2916,
+      "step": 756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.059814214706421,
+      "learning_rate": 4.999802548003039e-05,
+      "loss": 1.5868,
+      "step": 757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0494635105133057,
+      "learning_rate": 4.999802283676403e-05,
+      "loss": 1.5919,
+      "step": 758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8652894496917725,
+      "learning_rate": 4.9998020193497676e-05,
+      "loss": 1.4418,
+      "step": 759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3294434547424316,
+      "learning_rate": 4.999801755023131e-05,
+      "loss": 1.4395,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9837260246276855,
+      "learning_rate": 4.9998014906964956e-05,
+      "loss": 1.5766,
+      "step": 761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.069365978240967,
+      "learning_rate": 4.999801226369859e-05,
+      "loss": 1.612,
+      "step": 762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6400107145309448,
+      "learning_rate": 4.999800962043224e-05,
+      "loss": 1.4331,
+      "step": 763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5809969902038574,
+      "learning_rate": 4.9998006977165886e-05,
+      "loss": 1.5086,
+      "step": 764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3734482526779175,
+      "learning_rate": 4.999800433389952e-05,
+      "loss": 1.2924,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.117537021636963,
+      "learning_rate": 4.9998001690633166e-05,
+      "loss": 1.5265,
+      "step": 766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.991595983505249,
+      "learning_rate": 4.99979990473668e-05,
+      "loss": 1.5711,
+      "step": 767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1077380180358887,
+      "learning_rate": 4.999799640410045e-05,
+      "loss": 1.4487,
+      "step": 768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7712615728378296,
+      "learning_rate": 4.999799376083409e-05,
+      "loss": 1.4176,
+      "step": 769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3694961071014404,
+      "learning_rate": 4.999799111756773e-05,
+      "loss": 1.4156,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.06594717502594,
+      "learning_rate": 4.999798847430137e-05,
+      "loss": 1.9608,
+      "step": 771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8261525630950928,
+      "learning_rate": 4.999798583103501e-05,
+      "loss": 2.8173,
+      "step": 772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6964821815490723,
+      "learning_rate": 4.999798318776866e-05,
+      "loss": 1.4802,
+      "step": 773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6514617204666138,
+      "learning_rate": 4.99979805445023e-05,
+      "loss": 1.4861,
+      "step": 774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0943567752838135,
+      "learning_rate": 4.999797790123594e-05,
+      "loss": 1.3407,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6179749965667725,
+      "learning_rate": 4.999797525796958e-05,
+      "loss": 1.4259,
+      "step": 776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9393062591552734,
+      "learning_rate": 4.999797261470322e-05,
+      "loss": 1.5294,
+      "step": 777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2292369604110718,
+      "learning_rate": 4.9997969971436866e-05,
+      "loss": 1.7336,
+      "step": 778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0220770835876465,
+      "learning_rate": 4.999796732817051e-05,
+      "loss": 1.4831,
+      "step": 779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8180274963378906,
+      "learning_rate": 4.9997964684904146e-05,
+      "loss": 1.3659,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3631136417388916,
+      "learning_rate": 4.999796204163779e-05,
+      "loss": 1.3216,
+      "step": 781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5242695808410645,
+      "learning_rate": 4.999795939837143e-05,
+      "loss": 1.2928,
+      "step": 782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6234519481658936,
+      "learning_rate": 4.9997956755105076e-05,
+      "loss": 1.3354,
+      "step": 783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2397795915603638,
+      "learning_rate": 4.999795411183872e-05,
+      "loss": 1.3575,
+      "step": 784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.381130337715149,
+      "learning_rate": 4.9997951468572356e-05,
+      "loss": 1.5604,
+      "step": 785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9373488426208496,
+      "learning_rate": 4.9997948825306e-05,
+      "loss": 1.4622,
+      "step": 786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5031845569610596,
+      "learning_rate": 4.999794618203964e-05,
+      "loss": 1.612,
+      "step": 787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.509802222251892,
+      "learning_rate": 4.9997943538773286e-05,
+      "loss": 1.3246,
+      "step": 788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.294278860092163,
+      "learning_rate": 4.999794089550692e-05,
+      "loss": 1.6175,
+      "step": 789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9120171070098877,
+      "learning_rate": 4.9997938252240566e-05,
+      "loss": 1.2992,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.312208652496338,
+      "learning_rate": 4.999793560897421e-05,
+      "loss": 1.4413,
+      "step": 791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6987334489822388,
+      "learning_rate": 4.999793296570785e-05,
+      "loss": 1.4098,
+      "step": 792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.806592583656311,
+      "learning_rate": 4.9997930322441496e-05,
+      "loss": 1.2197,
+      "step": 793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7629849910736084,
+      "learning_rate": 4.999792767917513e-05,
+      "loss": 1.3978,
+      "step": 794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6859359741210938,
+      "learning_rate": 4.9997925035908775e-05,
+      "loss": 1.5791,
+      "step": 795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5475842952728271,
+      "learning_rate": 4.999792239264242e-05,
+      "loss": 1.3755,
+      "step": 796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3518673181533813,
+      "learning_rate": 4.999791974937606e-05,
+      "loss": 1.2961,
+      "step": 797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4943870306015015,
+      "learning_rate": 4.99979171061097e-05,
+      "loss": 1.3862,
+      "step": 798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5249680280685425,
+      "learning_rate": 4.999791446284334e-05,
+      "loss": 1.4053,
+      "step": 799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.051823139190674,
+      "learning_rate": 4.9997911819576985e-05,
+      "loss": 1.4297,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8780544996261597,
+      "learning_rate": 4.999790917631063e-05,
+      "loss": 1.509,
+      "step": 801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8331197500228882,
+      "learning_rate": 4.999790653304427e-05,
+      "loss": 1.662,
+      "step": 802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7669612169265747,
+      "learning_rate": 4.999790388977791e-05,
+      "loss": 1.3695,
+      "step": 803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5018846988677979,
+      "learning_rate": 4.999790124651155e-05,
+      "loss": 1.3652,
+      "step": 804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.39602530002594,
+      "learning_rate": 4.999789860324519e-05,
+      "loss": 1.3295,
+      "step": 805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.483156442642212,
+      "learning_rate": 4.999789595997884e-05,
+      "loss": 1.6529,
+      "step": 806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6420493125915527,
+      "learning_rate": 4.9997893316712475e-05,
+      "loss": 1.3601,
+      "step": 807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8679652214050293,
+      "learning_rate": 4.999789067344612e-05,
+      "loss": 1.616,
+      "step": 808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1648387908935547,
+      "learning_rate": 4.999788803017976e-05,
+      "loss": 1.5278,
+      "step": 809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7638260126113892,
+      "learning_rate": 4.99978853869134e-05,
+      "loss": 1.4052,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7524733543395996,
+      "learning_rate": 4.999788274364705e-05,
+      "loss": 1.3413,
+      "step": 811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.085515022277832,
+      "learning_rate": 4.9997880100380685e-05,
+      "loss": 1.5525,
+      "step": 812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4708447456359863,
+      "learning_rate": 4.999787745711433e-05,
+      "loss": 1.5609,
+      "step": 813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9660825729370117,
+      "learning_rate": 4.9997874813847965e-05,
+      "loss": 1.5597,
+      "step": 814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9223777055740356,
+      "learning_rate": 4.999787217058161e-05,
+      "loss": 1.4473,
+      "step": 815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7514933347702026,
+      "learning_rate": 4.999786952731526e-05,
+      "loss": 1.4402,
+      "step": 816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6003210544586182,
+      "learning_rate": 4.9997866884048895e-05,
+      "loss": 1.3511,
+      "step": 817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5094237327575684,
+      "learning_rate": 4.999786424078254e-05,
+      "loss": 1.5285,
+      "step": 818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7487809658050537,
+      "learning_rate": 4.9997861597516175e-05,
+      "loss": 1.6171,
+      "step": 819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0969631671905518,
+      "learning_rate": 4.999785895424982e-05,
+      "loss": 1.4138,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8949148654937744,
+      "learning_rate": 4.999785631098346e-05,
+      "loss": 1.3909,
+      "step": 821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9382295608520508,
+      "learning_rate": 4.9997853667717105e-05,
+      "loss": 1.5231,
+      "step": 822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9390904903411865,
+      "learning_rate": 4.999785102445074e-05,
+      "loss": 1.736,
+      "step": 823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4164990186691284,
+      "learning_rate": 4.9997848381184385e-05,
+      "loss": 1.3483,
+      "step": 824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3359124660491943,
+      "learning_rate": 4.999784573791803e-05,
+      "loss": 1.3575,
+      "step": 825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1327621936798096,
+      "learning_rate": 4.999784309465167e-05,
+      "loss": 1.5025,
+      "step": 826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3656070232391357,
+      "learning_rate": 4.9997840451385315e-05,
+      "loss": 1.3071,
+      "step": 827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8655412197113037,
+      "learning_rate": 4.999783780811895e-05,
+      "loss": 1.5446,
+      "step": 828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.796515703201294,
+      "learning_rate": 4.9997835164852595e-05,
+      "loss": 1.4262,
+      "step": 829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.750471591949463,
+      "learning_rate": 4.999783252158624e-05,
+      "loss": 1.3991,
+      "step": 830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7931029796600342,
+      "learning_rate": 4.999782987831988e-05,
+      "loss": 1.317,
+      "step": 831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5955181121826172,
+      "learning_rate": 4.999782723505352e-05,
+      "loss": 1.5577,
+      "step": 832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.70072603225708,
+      "learning_rate": 4.999782459178716e-05,
+      "loss": 1.2954,
+      "step": 833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9796757698059082,
+      "learning_rate": 4.9997821948520805e-05,
+      "loss": 1.5757,
+      "step": 834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3242762088775635,
+      "learning_rate": 4.999781930525445e-05,
+      "loss": 1.4087,
+      "step": 835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.149378776550293,
+      "learning_rate": 4.999781666198809e-05,
+      "loss": 1.4994,
+      "step": 836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.874853253364563,
+      "learning_rate": 4.999781401872173e-05,
+      "loss": 1.3935,
+      "step": 837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3929903507232666,
+      "learning_rate": 4.999781137545537e-05,
+      "loss": 1.6166,
+      "step": 838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.735087513923645,
+      "learning_rate": 4.999780873218901e-05,
+      "loss": 1.4336,
+      "step": 839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.783671259880066,
+      "learning_rate": 4.999780608892266e-05,
+      "loss": 1.4298,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5534485578536987,
+      "learning_rate": 4.9997803445656294e-05,
+      "loss": 1.3102,
+      "step": 841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.475335955619812,
+      "learning_rate": 4.999780080238994e-05,
+      "loss": 1.274,
+      "step": 842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5279887914657593,
+      "learning_rate": 4.999779815912358e-05,
+      "loss": 1.4263,
+      "step": 843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.537147879600525,
+      "learning_rate": 4.999779551585722e-05,
+      "loss": 1.386,
+      "step": 844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7355021238327026,
+      "learning_rate": 4.999779287259087e-05,
+      "loss": 1.5586,
+      "step": 845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.618196725845337,
+      "learning_rate": 4.9997790229324504e-05,
+      "loss": 1.3963,
+      "step": 846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5926192998886108,
+      "learning_rate": 4.999778758605815e-05,
+      "loss": 1.4351,
+      "step": 847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6407785415649414,
+      "learning_rate": 4.9997784942791784e-05,
+      "loss": 1.4247,
+      "step": 848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3287138938903809,
+      "learning_rate": 4.999778229952543e-05,
+      "loss": 1.3531,
+      "step": 849
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6459431648254395,
+      "learning_rate": 4.999777965625907e-05,
+      "loss": 1.3663,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4914357662200928,
+      "learning_rate": 4.9997777012992714e-05,
+      "loss": 1.4847,
+      "step": 851
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4533796310424805,
+      "learning_rate": 4.999777436972636e-05,
+      "loss": 1.4123,
+      "step": 852
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6796612739562988,
+      "learning_rate": 4.9997771726459994e-05,
+      "loss": 1.5008,
+      "step": 853
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4460468292236328,
+      "learning_rate": 4.999776908319364e-05,
+      "loss": 1.2839,
+      "step": 854
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3113656044006348,
+      "learning_rate": 4.999776643992728e-05,
+      "loss": 1.4592,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3787059783935547,
+      "learning_rate": 4.9997763796660924e-05,
+      "loss": 1.3721,
+      "step": 856
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8876813650131226,
+      "learning_rate": 4.999776115339456e-05,
+      "loss": 1.5553,
+      "step": 857
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.009164810180664,
+      "learning_rate": 4.9997758510128204e-05,
+      "loss": 1.4178,
+      "step": 858
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6356641054153442,
+      "learning_rate": 4.999775586686185e-05,
+      "loss": 1.4934,
+      "step": 859
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8387013673782349,
+      "learning_rate": 4.999775322359549e-05,
+      "loss": 1.2705,
+      "step": 860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5657459497451782,
+      "learning_rate": 4.9997750580329134e-05,
+      "loss": 1.4327,
+      "step": 861
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.600525379180908,
+      "learning_rate": 4.999774793706277e-05,
+      "loss": 1.7598,
+      "step": 862
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.887465000152588,
+      "learning_rate": 4.9997745293796414e-05,
+      "loss": 1.585,
+      "step": 863
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6064426898956299,
+      "learning_rate": 4.999774265053006e-05,
+      "loss": 2.6184,
+      "step": 864
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.143066167831421,
+      "learning_rate": 4.99977400072637e-05,
+      "loss": 1.3817,
+      "step": 865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8384140729904175,
+      "learning_rate": 4.999773736399734e-05,
+      "loss": 1.5036,
+      "step": 866
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2138593196868896,
+      "learning_rate": 4.999773472073098e-05,
+      "loss": 1.4316,
+      "step": 867
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8484307527542114,
+      "learning_rate": 4.9997732077464624e-05,
+      "loss": 1.5097,
+      "step": 868
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5860952138900757,
+      "learning_rate": 4.999772943419827e-05,
+      "loss": 1.5347,
+      "step": 869
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0189990997314453,
+      "learning_rate": 4.999772679093191e-05,
+      "loss": 1.5083,
+      "step": 870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5615359544754028,
+      "learning_rate": 4.999772414766555e-05,
+      "loss": 1.3805,
+      "step": 871
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8444371223449707,
+      "learning_rate": 4.999772150439919e-05,
+      "loss": 1.5305,
+      "step": 872
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.607749581336975,
+      "learning_rate": 4.9997718861132834e-05,
+      "loss": 1.509,
+      "step": 873
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.891762375831604,
+      "learning_rate": 4.999771621786648e-05,
+      "loss": 1.286,
+      "step": 874
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7921420335769653,
+      "learning_rate": 4.9997713574600114e-05,
+      "loss": 1.4324,
+      "step": 875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6742688417434692,
+      "learning_rate": 4.999771093133376e-05,
+      "loss": 1.5293,
+      "step": 876
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3771553039550781,
+      "learning_rate": 4.99977082880674e-05,
+      "loss": 1.4186,
+      "step": 877
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8694382905960083,
+      "learning_rate": 4.9997705644801044e-05,
+      "loss": 2.2662,
+      "step": 878
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7405458688735962,
+      "learning_rate": 4.999770300153469e-05,
+      "loss": 1.6285,
+      "step": 879
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.441399335861206,
+      "learning_rate": 4.9997700358268324e-05,
+      "loss": 1.5002,
+      "step": 880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3542197942733765,
+      "learning_rate": 4.999769771500197e-05,
+      "loss": 1.3549,
+      "step": 881
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6116493940353394,
+      "learning_rate": 4.9997695071735604e-05,
+      "loss": 1.3548,
+      "step": 882
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9453681707382202,
+      "learning_rate": 4.9997692428469254e-05,
+      "loss": 1.7188,
+      "step": 883
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8987818956375122,
+      "learning_rate": 4.999768978520289e-05,
+      "loss": 1.5013,
+      "step": 884
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0498034954071045,
+      "learning_rate": 4.9997687141936534e-05,
+      "loss": 1.3053,
+      "step": 885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.218341112136841,
+      "learning_rate": 4.999768449867018e-05,
+      "loss": 1.6216,
+      "step": 886
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0861144065856934,
+      "learning_rate": 4.9997681855403813e-05,
+      "loss": 2.3851,
+      "step": 887
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3664816617965698,
+      "learning_rate": 4.9997679212137464e-05,
+      "loss": 1.1511,
+      "step": 888
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8194330930709839,
+      "learning_rate": 4.99976765688711e-05,
+      "loss": 1.4831,
+      "step": 889
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.574248790740967,
+      "learning_rate": 4.9997673925604743e-05,
+      "loss": 1.5542,
+      "step": 890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.433059811592102,
+      "learning_rate": 4.999767128233838e-05,
+      "loss": 1.4093,
+      "step": 891
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6492595672607422,
+      "learning_rate": 4.999766863907202e-05,
+      "loss": 1.4024,
+      "step": 892
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.404905915260315,
+      "learning_rate": 4.999766599580567e-05,
+      "loss": 1.338,
+      "step": 893
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6576944589614868,
+      "learning_rate": 4.999766335253931e-05,
+      "loss": 1.3955,
+      "step": 894
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4513832330703735,
+      "learning_rate": 4.999766070927295e-05,
+      "loss": 1.3729,
+      "step": 895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4721410274505615,
+      "learning_rate": 4.999765806600659e-05,
+      "loss": 1.4123,
+      "step": 896
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5918679237365723,
+      "learning_rate": 4.999765542274023e-05,
+      "loss": 1.3868,
+      "step": 897
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5387637615203857,
+      "learning_rate": 4.999765277947388e-05,
+      "loss": 1.2347,
+      "step": 898
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8665485382080078,
+      "learning_rate": 4.999765013620752e-05,
+      "loss": 1.4712,
+      "step": 899
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5706149339675903,
+      "learning_rate": 4.9997647492941157e-05,
+      "loss": 1.3617,
+      "step": 900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5057519674301147,
+      "learning_rate": 4.99976448496748e-05,
+      "loss": 1.5074,
+      "step": 901
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.812976360321045,
+      "learning_rate": 4.999764220640844e-05,
+      "loss": 1.5763,
+      "step": 902
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.436614990234375,
+      "learning_rate": 4.9997639563142087e-05,
+      "loss": 1.4432,
+      "step": 903
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4688016176223755,
+      "learning_rate": 4.999763691987573e-05,
+      "loss": 1.4238,
+      "step": 904
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.487614393234253,
+      "learning_rate": 4.9997634276609366e-05,
+      "loss": 1.4172,
+      "step": 905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.029387950897217,
+      "learning_rate": 4.999763163334301e-05,
+      "loss": 1.2082,
+      "step": 906
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7056689262390137,
+      "learning_rate": 4.999762899007665e-05,
+      "loss": 1.4107,
+      "step": 907
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6955896615982056,
+      "learning_rate": 4.9997626346810296e-05,
+      "loss": 1.6001,
+      "step": 908
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8544981479644775,
+      "learning_rate": 4.999762370354393e-05,
+      "loss": 1.5063,
+      "step": 909
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9682278633117676,
+      "learning_rate": 4.9997621060277576e-05,
+      "loss": 1.7004,
+      "step": 910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3689627647399902,
+      "learning_rate": 4.999761841701121e-05,
+      "loss": 1.6017,
+      "step": 911
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8397393226623535,
+      "learning_rate": 4.999761577374486e-05,
+      "loss": 1.4972,
+      "step": 912
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8666011095046997,
+      "learning_rate": 4.9997613130478506e-05,
+      "loss": 1.6142,
+      "step": 913
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4697843790054321,
+      "learning_rate": 4.999761048721214e-05,
+      "loss": 1.4348,
+      "step": 914
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7959325313568115,
+      "learning_rate": 4.9997607843945786e-05,
+      "loss": 1.5606,
+      "step": 915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6873735189437866,
+      "learning_rate": 4.999760520067942e-05,
+      "loss": 1.4177,
+      "step": 916
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7464243173599243,
+      "learning_rate": 4.999760255741307e-05,
+      "loss": 1.4731,
+      "step": 917
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0482177734375,
+      "learning_rate": 4.999759991414671e-05,
+      "loss": 1.5142,
+      "step": 918
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5116355419158936,
+      "learning_rate": 4.999759727088035e-05,
+      "loss": 1.4698,
+      "step": 919
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2869035005569458,
+      "learning_rate": 4.9997594627613996e-05,
+      "loss": 1.2548,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2638843059539795,
+      "learning_rate": 4.999759198434763e-05,
+      "loss": 1.318,
+      "step": 921
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8976140022277832,
+      "learning_rate": 4.999758934108128e-05,
+      "loss": 1.4448,
+      "step": 922
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3364217281341553,
+      "learning_rate": 4.999758669781492e-05,
+      "loss": 1.3197,
+      "step": 923
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.274479627609253,
+      "learning_rate": 4.999758405454856e-05,
+      "loss": 1.4755,
+      "step": 924
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.204674482345581,
+      "learning_rate": 4.99975814112822e-05,
+      "loss": 1.655,
+      "step": 925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6114730834960938,
+      "learning_rate": 4.999757876801584e-05,
+      "loss": 1.3972,
+      "step": 926
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6629672050476074,
+      "learning_rate": 4.9997576124749486e-05,
+      "loss": 1.4506,
+      "step": 927
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.762386441230774,
+      "learning_rate": 4.999757348148313e-05,
+      "loss": 1.3748,
+      "step": 928
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5113979578018188,
+      "learning_rate": 4.999757083821677e-05,
+      "loss": 1.5638,
+      "step": 929
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2885488271713257,
+      "learning_rate": 4.999756819495041e-05,
+      "loss": 1.4336,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0407345294952393,
+      "learning_rate": 4.999756555168405e-05,
+      "loss": 1.6061,
+      "step": 931
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6766200065612793,
+      "learning_rate": 4.9997562908417696e-05,
+      "loss": 1.513,
+      "step": 932
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5082571506500244,
+      "learning_rate": 4.999756026515134e-05,
+      "loss": 1.2987,
+      "step": 933
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5236883163452148,
+      "learning_rate": 4.9997557621884976e-05,
+      "loss": 1.4709,
+      "step": 934
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.502068042755127,
+      "learning_rate": 4.999755497861862e-05,
+      "loss": 1.543,
+      "step": 935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.370849370956421,
+      "learning_rate": 4.999755233535226e-05,
+      "loss": 1.3066,
+      "step": 936
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1242148876190186,
+      "learning_rate": 4.9997549692085906e-05,
+      "loss": 1.4921,
+      "step": 937
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.52813720703125,
+      "learning_rate": 4.999754704881955e-05,
+      "loss": 1.3843,
+      "step": 938
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8818796873092651,
+      "learning_rate": 4.9997544405553186e-05,
+      "loss": 1.3557,
+      "step": 939
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.869711995124817,
+      "learning_rate": 4.999754176228683e-05,
+      "loss": 1.3001,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9590706825256348,
+      "learning_rate": 4.999753911902047e-05,
+      "loss": 1.3434,
+      "step": 941
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3641645908355713,
+      "learning_rate": 4.9997536475754116e-05,
+      "loss": 1.5457,
+      "step": 942
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5833516120910645,
+      "learning_rate": 4.999753383248775e-05,
+      "loss": 1.4165,
+      "step": 943
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.560587763786316,
+      "learning_rate": 4.9997531189221396e-05,
+      "loss": 1.2956,
+      "step": 944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3180065155029297,
+      "learning_rate": 4.999752854595504e-05,
+      "loss": 1.382,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6724289655685425,
+      "learning_rate": 4.999752590268868e-05,
+      "loss": 1.4993,
+      "step": 946
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7111526727676392,
+      "learning_rate": 4.9997523259422326e-05,
+      "loss": 1.5054,
+      "step": 947
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7522109746932983,
+      "learning_rate": 4.999752061615596e-05,
+      "loss": 1.5117,
+      "step": 948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5156182050704956,
+      "learning_rate": 4.9997517972889606e-05,
+      "loss": 1.6038,
+      "step": 949
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4357692003250122,
+      "learning_rate": 4.999751532962325e-05,
+      "loss": 1.438,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.668811559677124,
+      "learning_rate": 4.999751268635689e-05,
+      "loss": 1.5145,
+      "step": 951
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2899181842803955,
+      "learning_rate": 4.999751004309053e-05,
+      "loss": 1.4144,
+      "step": 952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3089768886566162,
+      "learning_rate": 4.999750739982417e-05,
+      "loss": 1.4062,
+      "step": 953
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6697150468826294,
+      "learning_rate": 4.999750475655781e-05,
+      "loss": 1.5143,
+      "step": 954
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8337292671203613,
+      "learning_rate": 4.999750211329146e-05,
+      "loss": 1.5275,
+      "step": 955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5023516416549683,
+      "learning_rate": 4.99974994700251e-05,
+      "loss": 1.5471,
+      "step": 956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4909359216690063,
+      "learning_rate": 4.999749682675874e-05,
+      "loss": 1.4216,
+      "step": 957
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.334818720817566,
+      "learning_rate": 4.999749418349238e-05,
+      "loss": 1.4258,
+      "step": 958
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.204435110092163,
+      "learning_rate": 4.999749154022602e-05,
+      "loss": 1.4537,
+      "step": 959
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4403231143951416,
+      "learning_rate": 4.999748889695967e-05,
+      "loss": 1.4969,
+      "step": 960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6100114583969116,
+      "learning_rate": 4.9997486253693305e-05,
+      "loss": 1.6031,
+      "step": 961
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4274693727493286,
+      "learning_rate": 4.999748361042695e-05,
+      "loss": 1.2966,
+      "step": 962
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7534230947494507,
+      "learning_rate": 4.9997480967160585e-05,
+      "loss": 1.4991,
+      "step": 963
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7229769229888916,
+      "learning_rate": 4.999747832389423e-05,
+      "loss": 1.4817,
+      "step": 964
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4905569553375244,
+      "learning_rate": 4.999747568062788e-05,
+      "loss": 1.4026,
+      "step": 965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6887929439544678,
+      "learning_rate": 4.9997473037361515e-05,
+      "loss": 1.5041,
+      "step": 966
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0461981296539307,
+      "learning_rate": 4.999747039409516e-05,
+      "loss": 1.5129,
+      "step": 967
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5247153043746948,
+      "learning_rate": 4.9997467750828795e-05,
+      "loss": 1.3858,
+      "step": 968
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.22855806350708,
+      "learning_rate": 4.999746510756244e-05,
+      "loss": 1.2118,
+      "step": 969
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6647753715515137,
+      "learning_rate": 4.999746246429608e-05,
+      "loss": 1.4647,
+      "step": 970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9681857824325562,
+      "learning_rate": 4.9997459821029725e-05,
+      "loss": 1.5634,
+      "step": 971
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4806772470474243,
+      "learning_rate": 4.999745717776337e-05,
+      "loss": 1.4919,
+      "step": 972
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.656294584274292,
+      "learning_rate": 4.9997454534497005e-05,
+      "loss": 1.4559,
+      "step": 973
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5756219625473022,
+      "learning_rate": 4.999745189123065e-05,
+      "loss": 1.3821,
+      "step": 974
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5561556816101074,
+      "learning_rate": 4.999744924796429e-05,
+      "loss": 1.4043,
+      "step": 975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4407517910003662,
+      "learning_rate": 4.9997446604697935e-05,
+      "loss": 1.5164,
+      "step": 976
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3915531635284424,
+      "learning_rate": 4.999744396143157e-05,
+      "loss": 1.37,
+      "step": 977
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2174906730651855,
+      "learning_rate": 4.9997441318165215e-05,
+      "loss": 1.6435,
+      "step": 978
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8146097660064697,
+      "learning_rate": 4.999743867489886e-05,
+      "loss": 1.4073,
+      "step": 979
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.209655284881592,
+      "learning_rate": 4.99974360316325e-05,
+      "loss": 1.6807,
+      "step": 980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3539373874664307,
+      "learning_rate": 4.9997433388366145e-05,
+      "loss": 1.5385,
+      "step": 981
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.178847074508667,
+      "learning_rate": 4.999743074509978e-05,
+      "loss": 1.5132,
+      "step": 982
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8641548156738281,
+      "learning_rate": 4.9997428101833425e-05,
+      "loss": 1.4191,
+      "step": 983
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8745845556259155,
+      "learning_rate": 4.999742545856707e-05,
+      "loss": 1.4598,
+      "step": 984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5757746696472168,
+      "learning_rate": 4.999742281530071e-05,
+      "loss": 1.5082,
+      "step": 985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.629859447479248,
+      "learning_rate": 4.999742017203435e-05,
+      "loss": 1.5637,
+      "step": 986
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0604467391967773,
+      "learning_rate": 4.999741752876799e-05,
+      "loss": 1.4609,
+      "step": 987
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7215290069580078,
+      "learning_rate": 4.999741488550163e-05,
+      "loss": 1.456,
+      "step": 988
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4035221338272095,
+      "learning_rate": 4.999741224223528e-05,
+      "loss": 1.3839,
+      "step": 989
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.249708652496338,
+      "learning_rate": 4.999740959896892e-05,
+      "loss": 1.6135,
+      "step": 990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6643279790878296,
+      "learning_rate": 4.999740695570256e-05,
+      "loss": 1.4623,
+      "step": 991
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.745256781578064,
+      "learning_rate": 4.99974043124362e-05,
+      "loss": 1.3079,
+      "step": 992
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9447695016860962,
+      "learning_rate": 4.999740166916984e-05,
+      "loss": 1.3986,
+      "step": 993
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8204765319824219,
+      "learning_rate": 4.999739902590349e-05,
+      "loss": 1.4973,
+      "step": 994
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.16280198097229,
+      "learning_rate": 4.9997396382637125e-05,
+      "loss": 1.7155,
+      "step": 995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6293866634368896,
+      "learning_rate": 4.999739373937077e-05,
+      "loss": 1.4665,
+      "step": 996
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.723814010620117,
+      "learning_rate": 4.9997391096104404e-05,
+      "loss": 1.731,
+      "step": 997
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6511163711547852,
+      "learning_rate": 4.999738845283805e-05,
+      "loss": 1.4547,
+      "step": 998
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0194947719573975,
+      "learning_rate": 4.99973858095717e-05,
+      "loss": 1.3818,
+      "step": 999
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.756489634513855,
+      "learning_rate": 4.9997383166305334e-05,
+      "loss": 1.4656,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9937820434570312,
+      "learning_rate": 4.999738052303898e-05,
+      "loss": 1.4376,
+      "step": 1001
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6054795980453491,
+      "learning_rate": 4.9997377879772614e-05,
+      "loss": 1.4612,
+      "step": 1002
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2201290130615234,
+      "learning_rate": 4.999737523650626e-05,
+      "loss": 3.0666,
+      "step": 1003
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.519354224205017,
+      "learning_rate": 4.99973725932399e-05,
+      "loss": 2.3767,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4099310636520386,
+      "learning_rate": 4.9997369949973544e-05,
+      "loss": 1.3569,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.518233299255371,
+      "learning_rate": 4.999736730670718e-05,
+      "loss": 1.3325,
+      "step": 1006
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2595713138580322,
+      "learning_rate": 4.9997364663440824e-05,
+      "loss": 1.4007,
+      "step": 1007
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6013710498809814,
+      "learning_rate": 4.999736202017447e-05,
+      "loss": 1.2563,
+      "step": 1008
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4424474239349365,
+      "learning_rate": 4.999735937690811e-05,
+      "loss": 1.2205,
+      "step": 1009
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9437572956085205,
+      "learning_rate": 4.9997356733641754e-05,
+      "loss": 1.5289,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.16794490814209,
+      "learning_rate": 4.999735409037539e-05,
+      "loss": 1.5451,
+      "step": 1011
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3731383085250854,
+      "learning_rate": 4.9997351447109034e-05,
+      "loss": 1.2641,
+      "step": 1012
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5051277875900269,
+      "learning_rate": 4.999734880384268e-05,
+      "loss": 1.4933,
+      "step": 1013
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.462992787361145,
+      "learning_rate": 4.999734616057632e-05,
+      "loss": 1.2261,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1178011894226074,
+      "learning_rate": 4.999734351730996e-05,
+      "loss": 1.6178,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0235447883605957,
+      "learning_rate": 4.99973408740436e-05,
+      "loss": 1.4737,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8876457214355469,
+      "learning_rate": 4.9997338230777244e-05,
+      "loss": 1.4937,
+      "step": 1017
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8709665536880493,
+      "learning_rate": 4.999733558751089e-05,
+      "loss": 2.3553,
+      "step": 1018
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2477513551712036,
+      "learning_rate": 4.999733294424453e-05,
+      "loss": 1.2433,
+      "step": 1019
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4810012578964233,
+      "learning_rate": 4.999733030097817e-05,
+      "loss": 1.4502,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.437100410461426,
+      "learning_rate": 4.999732765771181e-05,
+      "loss": 1.4176,
+      "step": 1021
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9280601739883423,
+      "learning_rate": 4.9997325014445454e-05,
+      "loss": 1.4322,
+      "step": 1022
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4793299436569214,
+      "learning_rate": 4.99973223711791e-05,
+      "loss": 1.4147,
+      "step": 1023
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7878303527832031,
+      "learning_rate": 4.999731972791274e-05,
+      "loss": 1.4955,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4679311513900757,
+      "learning_rate": 4.999731708464638e-05,
+      "loss": 1.3571,
+      "step": 1025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6806598901748657,
+      "learning_rate": 4.999731444138002e-05,
+      "loss": 1.6303,
+      "step": 1026
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6093957424163818,
+      "learning_rate": 4.9997311798113664e-05,
+      "loss": 1.3527,
+      "step": 1027
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.735363721847534,
+      "learning_rate": 4.999730915484731e-05,
+      "loss": 1.6592,
+      "step": 1028
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5004563331604004,
+      "learning_rate": 4.9997306511580944e-05,
+      "loss": 1.4548,
+      "step": 1029
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.518447756767273,
+      "learning_rate": 4.999730386831459e-05,
+      "loss": 1.3047,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2958728075027466,
+      "learning_rate": 4.9997301225048224e-05,
+      "loss": 1.3912,
+      "step": 1031
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8052082061767578,
+      "learning_rate": 4.9997298581781874e-05,
+      "loss": 1.4186,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6631898880004883,
+      "learning_rate": 4.999729593851552e-05,
+      "loss": 1.359,
+      "step": 1033
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8480231761932373,
+      "learning_rate": 4.9997293295249154e-05,
+      "loss": 1.3475,
+      "step": 1034
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.144474744796753,
+      "learning_rate": 4.99972906519828e-05,
+      "loss": 1.4923,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.403292179107666,
+      "learning_rate": 4.9997288008716434e-05,
+      "loss": 1.6539,
+      "step": 1036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5459891557693481,
+      "learning_rate": 4.9997285365450084e-05,
+      "loss": 1.4355,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.506888747215271,
+      "learning_rate": 4.999728272218372e-05,
+      "loss": 1.2662,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5370957851409912,
+      "learning_rate": 4.9997280078917364e-05,
+      "loss": 1.4925,
+      "step": 1039
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5832178592681885,
+      "learning_rate": 4.9997277435651e-05,
+      "loss": 1.3786,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3882402181625366,
+      "learning_rate": 4.9997274792384643e-05,
+      "loss": 1.4708,
+      "step": 1041
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3332628011703491,
+      "learning_rate": 4.9997272149118294e-05,
+      "loss": 1.2217,
+      "step": 1042
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7012501955032349,
+      "learning_rate": 4.999726950585193e-05,
+      "loss": 1.5434,
+      "step": 1043
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5218183994293213,
+      "learning_rate": 4.9997266862585574e-05,
+      "loss": 1.4641,
+      "step": 1044
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.296952247619629,
+      "learning_rate": 4.999726421931921e-05,
+      "loss": 1.172,
+      "step": 1045
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.70621919631958,
+      "learning_rate": 4.9997261576052853e-05,
+      "loss": 1.5794,
+      "step": 1046
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4940935373306274,
+      "learning_rate": 4.99972589327865e-05,
+      "loss": 1.3032,
+      "step": 1047
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0624289512634277,
+      "learning_rate": 4.999725628952014e-05,
+      "loss": 1.621,
+      "step": 1048
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7914116382598877,
+      "learning_rate": 4.999725364625378e-05,
+      "loss": 1.5053,
+      "step": 1049
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6048835515975952,
+      "learning_rate": 4.999725100298742e-05,
+      "loss": 1.4327,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6897687911987305,
+      "learning_rate": 4.999724835972106e-05,
+      "loss": 2.6294,
+      "step": 1051
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4562160968780518,
+      "learning_rate": 4.999724571645471e-05,
+      "loss": 1.2595,
+      "step": 1052
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8789695501327515,
+      "learning_rate": 4.999724307318835e-05,
+      "loss": 1.5282,
+      "step": 1053
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8595483303070068,
+      "learning_rate": 4.9997240429921987e-05,
+      "loss": 1.4774,
+      "step": 1054
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8779937028884888,
+      "learning_rate": 4.999723778665563e-05,
+      "loss": 1.3714,
+      "step": 1055
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.513359785079956,
+      "learning_rate": 4.999723514338927e-05,
+      "loss": 1.5295,
+      "step": 1056
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8019863367080688,
+      "learning_rate": 4.9997232500122917e-05,
+      "loss": 1.4936,
+      "step": 1057
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9174796342849731,
+      "learning_rate": 4.999722985685655e-05,
+      "loss": 1.3424,
+      "step": 1058
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4950934648513794,
+      "learning_rate": 4.9997227213590196e-05,
+      "loss": 1.4667,
+      "step": 1059
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8323005437850952,
+      "learning_rate": 4.999722457032384e-05,
+      "loss": 1.5498,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.588783025741577,
+      "learning_rate": 4.999722192705748e-05,
+      "loss": 1.5902,
+      "step": 1061
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7808047533035278,
+      "learning_rate": 4.9997219283791126e-05,
+      "loss": 1.6869,
+      "step": 1062
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.354308605194092,
+      "learning_rate": 4.999721664052476e-05,
+      "loss": 1.4444,
+      "step": 1063
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9849350452423096,
+      "learning_rate": 4.9997213997258406e-05,
+      "loss": 1.4125,
+      "step": 1064
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.830345869064331,
+      "learning_rate": 4.999721135399204e-05,
+      "loss": 1.5636,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5308046340942383,
+      "learning_rate": 4.999720871072569e-05,
+      "loss": 1.4912,
+      "step": 1066
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.959488034248352,
+      "learning_rate": 4.9997206067459336e-05,
+      "loss": 1.4591,
+      "step": 1067
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.676853060722351,
+      "learning_rate": 4.999720342419297e-05,
+      "loss": 1.3777,
+      "step": 1068
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.370880603790283,
+      "learning_rate": 4.9997200780926616e-05,
+      "loss": 1.4587,
+      "step": 1069
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.078446626663208,
+      "learning_rate": 4.999719813766025e-05,
+      "loss": 1.293,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2805958986282349,
+      "learning_rate": 4.99971954943939e-05,
+      "loss": 1.3789,
+      "step": 1071
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4300426244735718,
+      "learning_rate": 4.999719285112754e-05,
+      "loss": 1.4872,
+      "step": 1072
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8505345582962036,
+      "learning_rate": 4.999719020786118e-05,
+      "loss": 1.4836,
+      "step": 1073
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3578180074691772,
+      "learning_rate": 4.999718756459482e-05,
+      "loss": 1.4199,
+      "step": 1074
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4490983486175537,
+      "learning_rate": 4.999718492132846e-05,
+      "loss": 1.5089,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0934274196624756,
+      "learning_rate": 4.999718227806211e-05,
+      "loss": 1.4288,
+      "step": 1076
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0132052898406982,
+      "learning_rate": 4.999717963479575e-05,
+      "loss": 1.4532,
+      "step": 1077
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.807387351989746,
+      "learning_rate": 4.999717699152939e-05,
+      "loss": 1.4243,
+      "step": 1078
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7166149616241455,
+      "learning_rate": 4.999717434826303e-05,
+      "loss": 1.4341,
+      "step": 1079
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6899126768112183,
+      "learning_rate": 4.999717170499667e-05,
+      "loss": 1.451,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8622941970825195,
+      "learning_rate": 4.9997169061730316e-05,
+      "loss": 1.5364,
+      "step": 1081
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.276799440383911,
+      "learning_rate": 4.999716641846396e-05,
+      "loss": 1.4088,
+      "step": 1082
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5687377452850342,
+      "learning_rate": 4.9997163775197596e-05,
+      "loss": 1.3829,
+      "step": 1083
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1661384105682373,
+      "learning_rate": 4.999716113193124e-05,
+      "loss": 1.4895,
+      "step": 1084
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.864538311958313,
+      "learning_rate": 4.999715848866488e-05,
+      "loss": 1.4973,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3626623153686523,
+      "learning_rate": 4.9997155845398526e-05,
+      "loss": 1.2849,
+      "step": 1086
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6638743877410889,
+      "learning_rate": 4.999715320213217e-05,
+      "loss": 1.3575,
+      "step": 1087
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.076258659362793,
+      "learning_rate": 4.9997150558865806e-05,
+      "loss": 1.292,
+      "step": 1088
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2092177867889404,
+      "learning_rate": 4.999714791559945e-05,
+      "loss": 1.7787,
+      "step": 1089
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7218056917190552,
+      "learning_rate": 4.999714527233309e-05,
+      "loss": 1.4978,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4170880317687988,
+      "learning_rate": 4.9997142629066736e-05,
+      "loss": 1.379,
+      "step": 1091
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6725519895553589,
+      "learning_rate": 4.999713998580037e-05,
+      "loss": 1.4058,
+      "step": 1092
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1043007373809814,
+      "learning_rate": 4.9997137342534016e-05,
+      "loss": 1.6965,
+      "step": 1093
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9855257272720337,
+      "learning_rate": 4.999713469926766e-05,
+      "loss": 1.5176,
+      "step": 1094
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5882941484451294,
+      "learning_rate": 4.99971320560013e-05,
+      "loss": 1.5102,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3503485918045044,
+      "learning_rate": 4.9997129412734946e-05,
+      "loss": 1.384,
+      "step": 1096
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7196986675262451,
+      "learning_rate": 4.999712676946858e-05,
+      "loss": 1.5511,
+      "step": 1097
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2928115129470825,
+      "learning_rate": 4.9997124126202226e-05,
+      "loss": 1.2987,
+      "step": 1098
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4264644384384155,
+      "learning_rate": 4.999712148293587e-05,
+      "loss": 1.3705,
+      "step": 1099
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9456642866134644,
+      "learning_rate": 4.999711883966951e-05,
+      "loss": 1.3081,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8170098066329956,
+      "learning_rate": 4.999711619640315e-05,
+      "loss": 1.4133,
+      "step": 1101
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.458323359489441,
+      "learning_rate": 4.999711355313679e-05,
+      "loss": 1.4302,
+      "step": 1102
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.068587303161621,
+      "learning_rate": 4.9997110909870436e-05,
+      "loss": 1.5135,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4800782203674316,
+      "learning_rate": 4.999710826660408e-05,
+      "loss": 1.3679,
+      "step": 1104
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0889158248901367,
+      "learning_rate": 4.999710562333772e-05,
+      "loss": 1.5087,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8937705755233765,
+      "learning_rate": 4.999710298007136e-05,
+      "loss": 1.5163,
+      "step": 1106
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6502938270568848,
+      "learning_rate": 4.9997100336805e-05,
+      "loss": 1.5686,
+      "step": 1107
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5668184757232666,
+      "learning_rate": 4.999709769353864e-05,
+      "loss": 1.3329,
+      "step": 1108
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.126579523086548,
+      "learning_rate": 4.999709505027229e-05,
+      "loss": 1.7343,
+      "step": 1109
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.511728286743164,
+      "learning_rate": 4.9997092407005925e-05,
+      "loss": 2.0497,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2509863376617432,
+      "learning_rate": 4.999708976373957e-05,
+      "loss": 1.3993,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.276156187057495,
+      "learning_rate": 4.999708712047321e-05,
+      "loss": 1.4858,
+      "step": 1112
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.505746603012085,
+      "learning_rate": 4.999708447720685e-05,
+      "loss": 1.6819,
+      "step": 1113
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2393476963043213,
+      "learning_rate": 4.99970818339405e-05,
+      "loss": 1.4274,
+      "step": 1114
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2889941930770874,
+      "learning_rate": 4.9997079190674135e-05,
+      "loss": 1.2127,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9352214336395264,
+      "learning_rate": 4.999707654740778e-05,
+      "loss": 1.6271,
+      "step": 1116
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.394484043121338,
+      "learning_rate": 4.9997073904141415e-05,
+      "loss": 1.6068,
+      "step": 1117
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4264661073684692,
+      "learning_rate": 4.999707126087506e-05,
+      "loss": 1.4143,
+      "step": 1118
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.822289228439331,
+      "learning_rate": 4.999706861760871e-05,
+      "loss": 1.5117,
+      "step": 1119
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6675904989242554,
+      "learning_rate": 4.9997065974342345e-05,
+      "loss": 1.8262,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6898775100708008,
+      "learning_rate": 4.999706333107599e-05,
+      "loss": 1.2845,
+      "step": 1121
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6456935405731201,
+      "learning_rate": 4.9997060687809625e-05,
+      "loss": 1.4563,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8239011764526367,
+      "learning_rate": 4.999705804454327e-05,
+      "loss": 1.5447,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.387681484222412,
+      "learning_rate": 4.999705540127691e-05,
+      "loss": 1.6333,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.284294843673706,
+      "learning_rate": 4.9997052758010555e-05,
+      "loss": 1.5991,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.706307888031006,
+      "learning_rate": 4.999705011474419e-05,
+      "loss": 1.5757,
+      "step": 1126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1453118324279785,
+      "learning_rate": 4.9997047471477835e-05,
+      "loss": 1.5235,
+      "step": 1127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5314228534698486,
+      "learning_rate": 4.999704482821148e-05,
+      "loss": 1.3893,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1141867637634277,
+      "learning_rate": 4.999704218494512e-05,
+      "loss": 1.6299,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3065992593765259,
+      "learning_rate": 4.9997039541678765e-05,
+      "loss": 1.3899,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3816145658493042,
+      "learning_rate": 4.99970368984124e-05,
+      "loss": 1.1856,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5857349634170532,
+      "learning_rate": 4.9997034255146045e-05,
+      "loss": 1.4399,
+      "step": 1132
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6841405630111694,
+      "learning_rate": 4.999703161187969e-05,
+      "loss": 1.4832,
+      "step": 1133
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.143246650695801,
+      "learning_rate": 4.999702896861333e-05,
+      "loss": 1.3933,
+      "step": 1134
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.832563042640686,
+      "learning_rate": 4.999702632534697e-05,
+      "loss": 1.4996,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.98793363571167,
+      "learning_rate": 4.999702368208061e-05,
+      "loss": 1.5966,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.561803936958313,
+      "learning_rate": 4.9997021038814255e-05,
+      "loss": 1.3884,
+      "step": 1137
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7613340616226196,
+      "learning_rate": 4.99970183955479e-05,
+      "loss": 1.5078,
+      "step": 1138
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9511587619781494,
+      "learning_rate": 4.999701575228154e-05,
+      "loss": 1.5931,
+      "step": 1139
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3444315195083618,
+      "learning_rate": 4.999701310901518e-05,
+      "loss": 2.3755,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3303945064544678,
+      "learning_rate": 4.999701046574882e-05,
+      "loss": 1.4877,
+      "step": 1141
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3026316165924072,
+      "learning_rate": 4.999700782248246e-05,
+      "loss": 1.5005,
+      "step": 1142
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6391441822052,
+      "learning_rate": 4.999700517921611e-05,
+      "loss": 1.4961,
+      "step": 1143
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4551931619644165,
+      "learning_rate": 4.9997002535949745e-05,
+      "loss": 1.4179,
+      "step": 1144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5249067544937134,
+      "learning_rate": 4.999699989268339e-05,
+      "loss": 1.4767,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9374648332595825,
+      "learning_rate": 4.999699724941703e-05,
+      "loss": 1.5834,
+      "step": 1146
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5193065404891968,
+      "learning_rate": 4.999699460615067e-05,
+      "loss": 1.4009,
+      "step": 1147
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2290213108062744,
+      "learning_rate": 4.999699196288432e-05,
+      "loss": 1.4887,
+      "step": 1148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5334678888320923,
+      "learning_rate": 4.9996989319617955e-05,
+      "loss": 1.3812,
+      "step": 1149
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3511511087417603,
+      "learning_rate": 4.99969866763516e-05,
+      "loss": 1.3864,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.876132011413574,
+      "learning_rate": 4.9996984033085234e-05,
+      "loss": 1.574,
+      "step": 1151
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5641671419143677,
+      "learning_rate": 4.999698138981888e-05,
+      "loss": 1.428,
+      "step": 1152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.501058578491211,
+      "learning_rate": 4.999697874655252e-05,
+      "loss": 1.3774,
+      "step": 1153
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7195335626602173,
+      "learning_rate": 4.9996976103286164e-05,
+      "loss": 1.4593,
+      "step": 1154
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.418423056602478,
+      "learning_rate": 4.999697346001981e-05,
+      "loss": 1.4493,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.778815746307373,
+      "learning_rate": 4.9996970816753444e-05,
+      "loss": 1.3567,
+      "step": 1156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6739193201065063,
+      "learning_rate": 4.999696817348709e-05,
+      "loss": 1.3649,
+      "step": 1157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7072865962982178,
+      "learning_rate": 4.999696553022073e-05,
+      "loss": 1.4787,
+      "step": 1158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.528456687927246,
+      "learning_rate": 4.9996962886954374e-05,
+      "loss": 1.4188,
+      "step": 1159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.643436312675476,
+      "learning_rate": 4.999696024368801e-05,
+      "loss": 1.4461,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9961943626403809,
+      "learning_rate": 4.9996957600421654e-05,
+      "loss": 1.6465,
+      "step": 1161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0344698429107666,
+      "learning_rate": 4.99969549571553e-05,
+      "loss": 1.9693,
+      "step": 1162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.799443244934082,
+      "learning_rate": 4.999695231388894e-05,
+      "loss": 1.5009,
+      "step": 1163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5921785831451416,
+      "learning_rate": 4.9996949670622584e-05,
+      "loss": 1.4955,
+      "step": 1164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5499465465545654,
+      "learning_rate": 4.999694702735622e-05,
+      "loss": 1.3006,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6224418878555298,
+      "learning_rate": 4.9996944384089864e-05,
+      "loss": 1.4886,
+      "step": 1166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.338780164718628,
+      "learning_rate": 4.999694174082351e-05,
+      "loss": 2.0315,
+      "step": 1167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2358133792877197,
+      "learning_rate": 4.999693909755715e-05,
+      "loss": 1.6154,
+      "step": 1168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9980632066726685,
+      "learning_rate": 4.999693645429079e-05,
+      "loss": 1.4039,
+      "step": 1169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4976882934570312,
+      "learning_rate": 4.999693381102443e-05,
+      "loss": 1.3897,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7968237400054932,
+      "learning_rate": 4.9996931167758074e-05,
+      "loss": 1.5092,
+      "step": 1171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.257016897201538,
+      "learning_rate": 4.999692852449172e-05,
+      "loss": 1.5183,
+      "step": 1172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.259650230407715,
+      "learning_rate": 4.999692588122536e-05,
+      "loss": 2.2235,
+      "step": 1173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3782525062561035,
+      "learning_rate": 4.9996923237959e-05,
+      "loss": 1.439,
+      "step": 1174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9392964839935303,
+      "learning_rate": 4.999692059469264e-05,
+      "loss": 1.878,
+      "step": 1175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6974231004714966,
+      "learning_rate": 4.9996917951426284e-05,
+      "loss": 1.4366,
+      "step": 1176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2246670722961426,
+      "learning_rate": 4.999691530815993e-05,
+      "loss": 1.5964,
+      "step": 1177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7701430320739746,
+      "learning_rate": 4.9996912664893564e-05,
+      "loss": 1.5274,
+      "step": 1178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7217148542404175,
+      "learning_rate": 4.999691002162721e-05,
+      "loss": 1.5273,
+      "step": 1179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8743746280670166,
+      "learning_rate": 4.999690737836085e-05,
+      "loss": 1.4327,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7141356468200684,
+      "learning_rate": 4.9996904735094494e-05,
+      "loss": 1.5569,
+      "step": 1181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9053692817687988,
+      "learning_rate": 4.999690209182814e-05,
+      "loss": 1.5251,
+      "step": 1182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7444089651107788,
+      "learning_rate": 4.9996899448561774e-05,
+      "loss": 1.5232,
+      "step": 1183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3780622482299805,
+      "learning_rate": 4.999689680529542e-05,
+      "loss": 1.3389,
+      "step": 1184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3063881397247314,
+      "learning_rate": 4.9996894162029054e-05,
+      "loss": 1.3333,
+      "step": 1185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.604400634765625,
+      "learning_rate": 4.9996891518762704e-05,
+      "loss": 1.4867,
+      "step": 1186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.760846495628357,
+      "learning_rate": 4.999688887549634e-05,
+      "loss": 1.5254,
+      "step": 1187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6331040859222412,
+      "learning_rate": 4.9996886232229984e-05,
+      "loss": 1.5746,
+      "step": 1188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2875951528549194,
+      "learning_rate": 4.999688358896363e-05,
+      "loss": 1.3239,
+      "step": 1189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5540940761566162,
+      "learning_rate": 4.9996880945697264e-05,
+      "loss": 1.4067,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0547127723693848,
+      "learning_rate": 4.9996878302430914e-05,
+      "loss": 1.5951,
+      "step": 1191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7167798280715942,
+      "learning_rate": 4.999687565916455e-05,
+      "loss": 1.547,
+      "step": 1192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0185134410858154,
+      "learning_rate": 4.9996873015898194e-05,
+      "loss": 1.6401,
+      "step": 1193
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6730151176452637,
+      "learning_rate": 4.999687037263183e-05,
+      "loss": 1.4432,
+      "step": 1194
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5434881448745728,
+      "learning_rate": 4.9996867729365474e-05,
+      "loss": 1.4844,
+      "step": 1195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5214306116104126,
+      "learning_rate": 4.999686508609912e-05,
+      "loss": 1.4543,
+      "step": 1196
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5433967113494873,
+      "learning_rate": 4.999686244283276e-05,
+      "loss": 1.3919,
+      "step": 1197
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6802328824996948,
+      "learning_rate": 4.9996859799566404e-05,
+      "loss": 1.5145,
+      "step": 1198
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.114231824874878,
+      "learning_rate": 4.999685715630004e-05,
+      "loss": 1.6407,
+      "step": 1199
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9556931257247925,
+      "learning_rate": 4.9996854513033683e-05,
+      "loss": 1.5941,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8066049814224243,
+      "learning_rate": 4.999685186976733e-05,
+      "loss": 1.3354,
+      "step": 1201
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9053621292114258,
+      "learning_rate": 4.999684922650097e-05,
+      "loss": 1.398,
+      "step": 1202
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4387603998184204,
+      "learning_rate": 4.999684658323461e-05,
+      "loss": 1.4424,
+      "step": 1203
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.626471757888794,
+      "learning_rate": 4.999684393996825e-05,
+      "loss": 1.4613,
+      "step": 1204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7204896211624146,
+      "learning_rate": 4.999684129670189e-05,
+      "loss": 1.4197,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.465135097503662,
+      "learning_rate": 4.999683865343554e-05,
+      "loss": 1.4335,
+      "step": 1206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.721026062965393,
+      "learning_rate": 4.999683601016918e-05,
+      "loss": 2.5189,
+      "step": 1207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3551099300384521,
+      "learning_rate": 4.9996833366902817e-05,
+      "loss": 1.3815,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5245885848999023,
+      "learning_rate": 4.999683072363646e-05,
+      "loss": 1.2403,
+      "step": 1209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.386080265045166,
+      "learning_rate": 4.99968280803701e-05,
+      "loss": 1.456,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.383570909500122,
+      "learning_rate": 4.9996825437103747e-05,
+      "loss": 1.3723,
+      "step": 1211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8493497371673584,
+      "learning_rate": 4.999682279383738e-05,
+      "loss": 1.5235,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.44692063331604,
+      "learning_rate": 4.9996820150571026e-05,
+      "loss": 1.399,
+      "step": 1213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.421512484550476,
+      "learning_rate": 4.999681750730466e-05,
+      "loss": 1.2267,
+      "step": 1214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8010824918746948,
+      "learning_rate": 4.999681486403831e-05,
+      "loss": 1.6307,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0524637699127197,
+      "learning_rate": 4.9996812220771957e-05,
+      "loss": 1.438,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5172425508499146,
+      "learning_rate": 4.999680957750559e-05,
+      "loss": 1.422,
+      "step": 1217
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5418041944503784,
+      "learning_rate": 4.9996806934239236e-05,
+      "loss": 1.3032,
+      "step": 1218
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3450632095336914,
+      "learning_rate": 4.999680429097287e-05,
+      "loss": 1.4416,
+      "step": 1219
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0224690437316895,
+      "learning_rate": 4.999680164770652e-05,
+      "loss": 1.5142,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4633586406707764,
+      "learning_rate": 4.999679900444016e-05,
+      "loss": 1.4324,
+      "step": 1221
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.433762311935425,
+      "learning_rate": 4.99967963611738e-05,
+      "loss": 1.5033,
+      "step": 1222
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.471069097518921,
+      "learning_rate": 4.9996793717907446e-05,
+      "loss": 1.4403,
+      "step": 1223
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5681265592575073,
+      "learning_rate": 4.999679107464108e-05,
+      "loss": 1.3979,
+      "step": 1224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6580241918563843,
+      "learning_rate": 4.999678843137473e-05,
+      "loss": 1.5576,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5420479774475098,
+      "learning_rate": 4.999678578810837e-05,
+      "loss": 1.4089,
+      "step": 1226
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9444007873535156,
+      "learning_rate": 4.999678314484201e-05,
+      "loss": 1.482,
+      "step": 1227
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.998642921447754,
+      "learning_rate": 4.999678050157565e-05,
+      "loss": 1.7195,
+      "step": 1228
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8256584405899048,
+      "learning_rate": 4.999677785830929e-05,
+      "loss": 1.404,
+      "step": 1229
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1949007511138916,
+      "learning_rate": 4.9996775215042936e-05,
+      "loss": 1.5984,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8621702194213867,
+      "learning_rate": 4.999677257177658e-05,
+      "loss": 1.4748,
+      "step": 1231
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6007260084152222,
+      "learning_rate": 4.999676992851022e-05,
+      "loss": 1.4584,
+      "step": 1232
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1717866659164429,
+      "learning_rate": 4.999676728524386e-05,
+      "loss": 2.1936,
+      "step": 1233
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.713120937347412,
+      "learning_rate": 4.99967646419775e-05,
+      "loss": 1.317,
+      "step": 1234
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5664331912994385,
+      "learning_rate": 4.9996761998711146e-05,
+      "loss": 1.3796,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6552459001541138,
+      "learning_rate": 4.999675935544479e-05,
+      "loss": 1.5103,
+      "step": 1236
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.339721441268921,
+      "learning_rate": 4.9996756712178426e-05,
+      "loss": 1.4115,
+      "step": 1237
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1506664752960205,
+      "learning_rate": 4.999675406891207e-05,
+      "loss": 1.4921,
+      "step": 1238
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0551700592041016,
+      "learning_rate": 4.999675142564571e-05,
+      "loss": 1.6769,
+      "step": 1239
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7394362688064575,
+      "learning_rate": 4.9996748782379356e-05,
+      "loss": 1.5202,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5325084924697876,
+      "learning_rate": 4.9996746139113e-05,
+      "loss": 1.6659,
+      "step": 1241
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1849703788757324,
+      "learning_rate": 4.9996743495846636e-05,
+      "loss": 1.3332,
+      "step": 1242
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2361855506896973,
+      "learning_rate": 4.999674085258028e-05,
+      "loss": 1.3663,
+      "step": 1243
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.106640338897705,
+      "learning_rate": 4.999673820931392e-05,
+      "loss": 1.4112,
+      "step": 1244
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.399346113204956,
+      "learning_rate": 4.9996735566047566e-05,
+      "loss": 1.4337,
+      "step": 1245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7224688529968262,
+      "learning_rate": 4.99967329227812e-05,
+      "loss": 1.4007,
+      "step": 1246
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8059715032577515,
+      "learning_rate": 4.9996730279514846e-05,
+      "loss": 1.479,
+      "step": 1247
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2955958843231201,
+      "learning_rate": 4.999672763624849e-05,
+      "loss": 1.3888,
+      "step": 1248
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3170642852783203,
+      "learning_rate": 4.999672499298213e-05,
+      "loss": 2.4594,
+      "step": 1249
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3453547954559326,
+      "learning_rate": 4.9996722349715776e-05,
+      "loss": 1.4604,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6591875553131104,
+      "learning_rate": 4.999671970644941e-05,
+      "loss": 1.444,
+      "step": 1251
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6264104843139648,
+      "learning_rate": 4.9996717063183056e-05,
+      "loss": 1.2394,
+      "step": 1252
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0645556449890137,
+      "learning_rate": 4.99967144199167e-05,
+      "loss": 1.4998,
+      "step": 1253
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0383880138397217,
+      "learning_rate": 4.999671177665034e-05,
+      "loss": 1.4742,
+      "step": 1254
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.218766212463379,
+      "learning_rate": 4.999670913338398e-05,
+      "loss": 1.4985,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7681058645248413,
+      "learning_rate": 4.999670649011762e-05,
+      "loss": 1.5162,
+      "step": 1256
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6157610416412354,
+      "learning_rate": 4.999670384685126e-05,
+      "loss": 1.4853,
+      "step": 1257
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.035855293273926,
+      "learning_rate": 4.999670120358491e-05,
+      "loss": 1.5423,
+      "step": 1258
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6555118560791016,
+      "learning_rate": 4.999669856031855e-05,
+      "loss": 1.3842,
+      "step": 1259
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1235692501068115,
+      "learning_rate": 4.999669591705219e-05,
+      "loss": 2.2904,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7495386600494385,
+      "learning_rate": 4.999669327378583e-05,
+      "loss": 1.4013,
+      "step": 1261
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4186112880706787,
+      "learning_rate": 4.999669063051947e-05,
+      "loss": 1.256,
+      "step": 1262
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.797945261001587,
+      "learning_rate": 4.999668798725312e-05,
+      "loss": 1.547,
+      "step": 1263
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4211623668670654,
+      "learning_rate": 4.9996685343986755e-05,
+      "loss": 1.3108,
+      "step": 1264
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.665160894393921,
+      "learning_rate": 4.99966827007204e-05,
+      "loss": 1.5,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.69222092628479,
+      "learning_rate": 4.9996680057454035e-05,
+      "loss": 1.7182,
+      "step": 1266
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.12614107131958,
+      "learning_rate": 4.999667741418768e-05,
+      "loss": 2.7384,
+      "step": 1267
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6549197435379028,
+      "learning_rate": 4.999667477092133e-05,
+      "loss": 1.5527,
+      "step": 1268
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.450705885887146,
+      "learning_rate": 4.9996672127654965e-05,
+      "loss": 1.831,
+      "step": 1269
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6073330640792847,
+      "learning_rate": 4.999666948438861e-05,
+      "loss": 1.433,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4506915807724,
+      "learning_rate": 4.9996666841122245e-05,
+      "loss": 1.301,
+      "step": 1271
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5855283737182617,
+      "learning_rate": 4.999666419785589e-05,
+      "loss": 1.5536,
+      "step": 1272
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1054000854492188,
+      "learning_rate": 4.999666155458953e-05,
+      "loss": 1.2321,
+      "step": 1273
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2472100257873535,
+      "learning_rate": 4.9996658911323175e-05,
+      "loss": 1.3049,
+      "step": 1274
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8185840845108032,
+      "learning_rate": 4.999665626805682e-05,
+      "loss": 1.5309,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0070900917053223,
+      "learning_rate": 4.9996653624790455e-05,
+      "loss": 1.4927,
+      "step": 1276
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6464122533798218,
+      "learning_rate": 4.99966509815241e-05,
+      "loss": 1.4327,
+      "step": 1277
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6889841556549072,
+      "learning_rate": 4.999664833825774e-05,
+      "loss": 1.3865,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6239495277404785,
+      "learning_rate": 4.9996645694991385e-05,
+      "loss": 1.5422,
+      "step": 1279
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8880192041397095,
+      "learning_rate": 4.999664305172502e-05,
+      "loss": 1.7577,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0377233028411865,
+      "learning_rate": 4.9996640408458665e-05,
+      "loss": 1.6262,
+      "step": 1281
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3179566860198975,
+      "learning_rate": 4.999663776519231e-05,
+      "loss": 1.5613,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9834368228912354,
+      "learning_rate": 4.999663512192595e-05,
+      "loss": 1.4782,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7646499872207642,
+      "learning_rate": 4.9996632478659595e-05,
+      "loss": 1.4332,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.303793430328369,
+      "learning_rate": 4.999662983539323e-05,
+      "loss": 1.4781,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5105208158493042,
+      "learning_rate": 4.9996627192126875e-05,
+      "loss": 1.3792,
+      "step": 1286
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.691324234008789,
+      "learning_rate": 4.999662454886052e-05,
+      "loss": 1.383,
+      "step": 1287
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.355541706085205,
+      "learning_rate": 4.999662190559416e-05,
+      "loss": 1.3941,
+      "step": 1288
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6435915231704712,
+      "learning_rate": 4.99966192623278e-05,
+      "loss": 1.4918,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.906650424003601,
+      "learning_rate": 4.999661661906144e-05,
+      "loss": 1.4146,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.167811155319214,
+      "learning_rate": 4.999661397579508e-05,
+      "loss": 1.6211,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4526634216308594,
+      "learning_rate": 4.999661133252873e-05,
+      "loss": 1.5229,
+      "step": 1292
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6625903844833374,
+      "learning_rate": 4.999660868926237e-05,
+      "loss": 1.3262,
+      "step": 1293
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5997741222381592,
+      "learning_rate": 4.999660604599601e-05,
+      "loss": 1.5165,
+      "step": 1294
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.771795392036438,
+      "learning_rate": 4.999660340272965e-05,
+      "loss": 1.4962,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.602679967880249,
+      "learning_rate": 4.999660075946329e-05,
+      "loss": 2.7217,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3461968898773193,
+      "learning_rate": 4.999659811619694e-05,
+      "loss": 1.4647,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6776432991027832,
+      "learning_rate": 4.9996595472930575e-05,
+      "loss": 1.4103,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6535465717315674,
+      "learning_rate": 4.999659282966422e-05,
+      "loss": 1.4829,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6408205032348633,
+      "learning_rate": 4.9996590186397855e-05,
+      "loss": 1.438,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8082473278045654,
+      "learning_rate": 4.99965875431315e-05,
+      "loss": 1.4596,
+      "step": 1301
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.222867012023926,
+      "learning_rate": 4.999658489986515e-05,
+      "loss": 1.4747,
+      "step": 1302
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7803757190704346,
+      "learning_rate": 4.9996582256598785e-05,
+      "loss": 1.5873,
+      "step": 1303
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8520716428756714,
+      "learning_rate": 4.999657961333243e-05,
+      "loss": 1.4662,
+      "step": 1304
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6587796211242676,
+      "learning_rate": 4.9996576970066064e-05,
+      "loss": 1.481,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6213680505752563,
+      "learning_rate": 4.999657432679971e-05,
+      "loss": 1.4731,
+      "step": 1306
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.745964527130127,
+      "learning_rate": 4.999657168353335e-05,
+      "loss": 1.6335,
+      "step": 1307
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.923446774482727,
+      "learning_rate": 4.9996569040266994e-05,
+      "loss": 1.6028,
+      "step": 1308
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3263885974884033,
+      "learning_rate": 4.999656639700063e-05,
+      "loss": 2.0213,
+      "step": 1309
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9593790769577026,
+      "learning_rate": 4.9996563753734274e-05,
+      "loss": 1.4151,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4535249471664429,
+      "learning_rate": 4.999656111046792e-05,
+      "loss": 1.4506,
+      "step": 1311
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6823080778121948,
+      "learning_rate": 4.999655846720156e-05,
+      "loss": 1.5267,
+      "step": 1312
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4022363424301147,
+      "learning_rate": 4.9996555823935204e-05,
+      "loss": 1.3215,
+      "step": 1313
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8521122932434082,
+      "learning_rate": 4.999655318066884e-05,
+      "loss": 1.4329,
+      "step": 1314
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.749710202217102,
+      "learning_rate": 4.9996550537402484e-05,
+      "loss": 1.48,
+      "step": 1315
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4217243194580078,
+      "learning_rate": 4.999654789413613e-05,
+      "loss": 1.3206,
+      "step": 1316
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3375098705291748,
+      "learning_rate": 4.999654525086977e-05,
+      "loss": 1.3316,
+      "step": 1317
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7874835729599,
+      "learning_rate": 4.999654260760341e-05,
+      "loss": 1.4801,
+      "step": 1318
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6397221088409424,
+      "learning_rate": 4.999653996433705e-05,
+      "loss": 1.451,
+      "step": 1319
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5537410974502563,
+      "learning_rate": 4.9996537321070694e-05,
+      "loss": 1.3495,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6712428331375122,
+      "learning_rate": 4.999653467780434e-05,
+      "loss": 1.9794,
+      "step": 1321
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8719918727874756,
+      "learning_rate": 4.999653203453798e-05,
+      "loss": 1.4898,
+      "step": 1322
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0124006271362305,
+      "learning_rate": 4.999652939127162e-05,
+      "loss": 1.5135,
+      "step": 1323
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7917810678482056,
+      "learning_rate": 4.999652674800526e-05,
+      "loss": 1.5069,
+      "step": 1324
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7538704872131348,
+      "learning_rate": 4.9996524104738904e-05,
+      "loss": 1.3661,
+      "step": 1325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.638077735900879,
+      "learning_rate": 4.999652146147255e-05,
+      "loss": 1.3896,
+      "step": 1326
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6370396614074707,
+      "learning_rate": 4.999651881820619e-05,
+      "loss": 1.479,
+      "step": 1327
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.569029688835144,
+      "learning_rate": 4.999651617493983e-05,
+      "loss": 1.4595,
+      "step": 1328
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8491169214248657,
+      "learning_rate": 4.999651353167347e-05,
+      "loss": 1.4964,
+      "step": 1329
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7683674097061157,
+      "learning_rate": 4.9996510888407114e-05,
+      "loss": 1.7308,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7877758741378784,
+      "learning_rate": 4.999650824514076e-05,
+      "loss": 1.2402,
+      "step": 1331
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.233752489089966,
+      "learning_rate": 4.9996505601874394e-05,
+      "loss": 1.6003,
+      "step": 1332
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5794938802719116,
+      "learning_rate": 4.999650295860804e-05,
+      "loss": 1.4141,
+      "step": 1333
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.585939645767212,
+      "learning_rate": 4.9996500315341674e-05,
+      "loss": 1.4742,
+      "step": 1334
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4793351888656616,
+      "learning_rate": 4.9996497672075324e-05,
+      "loss": 1.4545,
+      "step": 1335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5079694986343384,
+      "learning_rate": 4.999649502880897e-05,
+      "loss": 1.4127,
+      "step": 1336
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2044994831085205,
+      "learning_rate": 4.9996492385542604e-05,
+      "loss": 1.4054,
+      "step": 1337
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.588876485824585,
+      "learning_rate": 4.999648974227625e-05,
+      "loss": 1.4909,
+      "step": 1338
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5806549787521362,
+      "learning_rate": 4.9996487099009884e-05,
+      "loss": 1.2126,
+      "step": 1339
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5384515523910522,
+      "learning_rate": 4.9996484455743534e-05,
+      "loss": 1.4401,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9238052368164062,
+      "learning_rate": 4.999648181247717e-05,
+      "loss": 1.5017,
+      "step": 1341
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6442115306854248,
+      "learning_rate": 4.9996479169210814e-05,
+      "loss": 1.4524,
+      "step": 1342
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8088775873184204,
+      "learning_rate": 4.999647652594445e-05,
+      "loss": 1.5632,
+      "step": 1343
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.016383647918701,
+      "learning_rate": 4.9996473882678094e-05,
+      "loss": 1.6216,
+      "step": 1344
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3727471828460693,
+      "learning_rate": 4.9996471239411744e-05,
+      "loss": 1.2304,
+      "step": 1345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3951318264007568,
+      "learning_rate": 4.999646859614538e-05,
+      "loss": 1.3689,
+      "step": 1346
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5403143167495728,
+      "learning_rate": 4.9996465952879024e-05,
+      "loss": 1.438,
+      "step": 1347
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9013888835906982,
+      "learning_rate": 4.999646330961266e-05,
+      "loss": 1.4725,
+      "step": 1348
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.753565788269043,
+      "learning_rate": 4.9996460666346304e-05,
+      "loss": 1.2725,
+      "step": 1349
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9699127674102783,
+      "learning_rate": 4.999645802307995e-05,
+      "loss": 1.5185,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9630236625671387,
+      "learning_rate": 4.999645537981359e-05,
+      "loss": 1.5121,
+      "step": 1351
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7425123453140259,
+      "learning_rate": 4.999645273654723e-05,
+      "loss": 1.5989,
+      "step": 1352
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.986499547958374,
+      "learning_rate": 4.999645009328087e-05,
+      "loss": 1.7677,
+      "step": 1353
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.100841522216797,
+      "learning_rate": 4.9996447450014513e-05,
+      "loss": 1.5681,
+      "step": 1354
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4215673208236694,
+      "learning_rate": 4.999644480674816e-05,
+      "loss": 1.3745,
+      "step": 1355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7875561714172363,
+      "learning_rate": 4.99964421634818e-05,
+      "loss": 1.5388,
+      "step": 1356
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7609808444976807,
+      "learning_rate": 4.999643952021544e-05,
+      "loss": 1.3811,
+      "step": 1357
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3373780250549316,
+      "learning_rate": 4.999643687694908e-05,
+      "loss": 1.5471,
+      "step": 1358
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6433287858963013,
+      "learning_rate": 4.999643423368272e-05,
+      "loss": 1.4494,
+      "step": 1359
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9269837141036987,
+      "learning_rate": 4.999643159041637e-05,
+      "loss": 1.4964,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7865773439407349,
+      "learning_rate": 4.999642894715e-05,
+      "loss": 1.4823,
+      "step": 1361
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3367929458618164,
+      "learning_rate": 4.9996426303883647e-05,
+      "loss": 1.4095,
+      "step": 1362
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5508781671524048,
+      "learning_rate": 4.999642366061729e-05,
+      "loss": 1.3951,
+      "step": 1363
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4639137983322144,
+      "learning_rate": 4.999642101735093e-05,
+      "loss": 1.4103,
+      "step": 1364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5176692008972168,
+      "learning_rate": 4.999641837408458e-05,
+      "loss": 1.4106,
+      "step": 1365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5664640665054321,
+      "learning_rate": 4.999641573081821e-05,
+      "loss": 1.5156,
+      "step": 1366
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6749517917633057,
+      "learning_rate": 4.9996413087551857e-05,
+      "loss": 1.673,
+      "step": 1367
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5534433126449585,
+      "learning_rate": 4.999641044428549e-05,
+      "loss": 1.4759,
+      "step": 1368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.892157793045044,
+      "learning_rate": 4.999640780101914e-05,
+      "loss": 1.4827,
+      "step": 1369
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4595266580581665,
+      "learning_rate": 4.999640515775278e-05,
+      "loss": 1.4482,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4922590255737305,
+      "learning_rate": 4.999640251448642e-05,
+      "loss": 1.4855,
+      "step": 1371
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5745081901550293,
+      "learning_rate": 4.9996399871220066e-05,
+      "loss": 1.4449,
+      "step": 1372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1136226654052734,
+      "learning_rate": 4.99963972279537e-05,
+      "loss": 1.6805,
+      "step": 1373
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6276012659072876,
+      "learning_rate": 4.999639458468735e-05,
+      "loss": 1.3723,
+      "step": 1374
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6983731985092163,
+      "learning_rate": 4.999639194142099e-05,
+      "loss": 1.4876,
+      "step": 1375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3539111614227295,
+      "learning_rate": 4.999638929815463e-05,
+      "loss": 1.2751,
+      "step": 1376
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.153937339782715,
+      "learning_rate": 4.999638665488827e-05,
+      "loss": 1.5818,
+      "step": 1377
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6860653162002563,
+      "learning_rate": 4.999638401162191e-05,
+      "loss": 1.4312,
+      "step": 1378
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0597805976867676,
+      "learning_rate": 4.999638136835556e-05,
+      "loss": 1.3506,
+      "step": 1379
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5780367851257324,
+      "learning_rate": 4.99963787250892e-05,
+      "loss": 1.4846,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6771891117095947,
+      "learning_rate": 4.999637608182284e-05,
+      "loss": 1.3074,
+      "step": 1381
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9793637990951538,
+      "learning_rate": 4.999637343855648e-05,
+      "loss": 1.5142,
+      "step": 1382
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4979267120361328,
+      "learning_rate": 4.999637079529012e-05,
+      "loss": 1.4303,
+      "step": 1383
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1569325923919678,
+      "learning_rate": 4.9996368152023766e-05,
+      "loss": 1.4829,
+      "step": 1384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4417877197265625,
+      "learning_rate": 4.999636550875741e-05,
+      "loss": 1.4953,
+      "step": 1385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5484139919281006,
+      "learning_rate": 4.9996362865491046e-05,
+      "loss": 1.5004,
+      "step": 1386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7802232503890991,
+      "learning_rate": 4.999636022222469e-05,
+      "loss": 1.4982,
+      "step": 1387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.101252555847168,
+      "learning_rate": 4.999635757895833e-05,
+      "loss": 1.5178,
+      "step": 1388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.152898073196411,
+      "learning_rate": 4.9996354935691976e-05,
+      "loss": 2.7456,
+      "step": 1389
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9095869064331055,
+      "learning_rate": 4.999635229242562e-05,
+      "loss": 1.5343,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6063610315322876,
+      "learning_rate": 4.9996349649159256e-05,
+      "loss": 1.3851,
+      "step": 1391
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2904754877090454,
+      "learning_rate": 4.99963470058929e-05,
+      "loss": 1.5138,
+      "step": 1392
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.578849196434021,
+      "learning_rate": 4.999634436262654e-05,
+      "loss": 1.3395,
+      "step": 1393
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6599199771881104,
+      "learning_rate": 4.9996341719360186e-05,
+      "loss": 1.6512,
+      "step": 1394
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.80825674533844,
+      "learning_rate": 4.999633907609382e-05,
+      "loss": 1.5892,
+      "step": 1395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5037869215011597,
+      "learning_rate": 4.9996336432827466e-05,
+      "loss": 2.2468,
+      "step": 1396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.716630220413208,
+      "learning_rate": 4.999633378956111e-05,
+      "loss": 1.532,
+      "step": 1397
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5493183135986328,
+      "learning_rate": 4.999633114629475e-05,
+      "loss": 1.4467,
+      "step": 1398
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5337332487106323,
+      "learning_rate": 4.9996328503028396e-05,
+      "loss": 1.2895,
+      "step": 1399
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6296724081039429,
+      "learning_rate": 4.999632585976203e-05,
+      "loss": 1.4735,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7963764667510986,
+      "learning_rate": 4.963230240549828e-05,
+      "loss": 1.3688,
+      "step": 1401
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.317366361618042,
+      "learning_rate": 4.9632038065027755e-05,
+      "loss": 1.6316,
+      "step": 1402
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4281914234161377,
+      "learning_rate": 4.9631773724557236e-05,
+      "loss": 1.5324,
+      "step": 1403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8440171480178833,
+      "learning_rate": 4.96315093840867e-05,
+      "loss": 1.5052,
+      "step": 1404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2446162700653076,
+      "learning_rate": 4.963124504361618e-05,
+      "loss": 1.5079,
+      "step": 1405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4120769500732422,
+      "learning_rate": 4.963098070314565e-05,
+      "loss": 1.4153,
+      "step": 1406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6582725048065186,
+      "learning_rate": 4.963071636267513e-05,
+      "loss": 1.4361,
+      "step": 1407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8487279415130615,
+      "learning_rate": 4.9630452022204603e-05,
+      "loss": 1.6114,
+      "step": 1408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6884549856185913,
+      "learning_rate": 4.963018768173408e-05,
+      "loss": 1.4886,
+      "step": 1409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5356805324554443,
+      "learning_rate": 4.962992334126355e-05,
+      "loss": 1.5441,
+      "step": 1410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.275338888168335,
+      "learning_rate": 4.9629659000793024e-05,
+      "loss": 1.3754,
+      "step": 1411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6018850803375244,
+      "learning_rate": 4.96293946603225e-05,
+      "loss": 1.4368,
+      "step": 1412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9272229671478271,
+      "learning_rate": 4.962913031985197e-05,
+      "loss": 1.4696,
+      "step": 1413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6697336435317993,
+      "learning_rate": 4.9628865979381445e-05,
+      "loss": 1.4542,
+      "step": 1414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.275710105895996,
+      "learning_rate": 4.962860163891092e-05,
+      "loss": 1.6617,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9890187978744507,
+      "learning_rate": 4.962833729844039e-05,
+      "loss": 1.3874,
+      "step": 1416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.525164246559143,
+      "learning_rate": 4.9628072957969866e-05,
+      "loss": 1.4283,
+      "step": 1417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8081562519073486,
+      "learning_rate": 4.9627808617499346e-05,
+      "loss": 1.344,
+      "step": 1418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5670658349990845,
+      "learning_rate": 4.962754427702881e-05,
+      "loss": 1.5209,
+      "step": 1419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6367779970169067,
+      "learning_rate": 4.962727993655829e-05,
+      "loss": 1.474,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4215291738510132,
+      "learning_rate": 4.962701559608776e-05,
+      "loss": 1.3514,
+      "step": 1421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9531952142715454,
+      "learning_rate": 4.962675125561724e-05,
+      "loss": 1.4367,
+      "step": 1422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6882989406585693,
+      "learning_rate": 4.9626486915146714e-05,
+      "loss": 1.2032,
+      "step": 1423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7947328090667725,
+      "learning_rate": 4.962622257467619e-05,
+      "loss": 1.4811,
+      "step": 1424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4503611326217651,
+      "learning_rate": 4.962595823420566e-05,
+      "loss": 1.4353,
+      "step": 1425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6771659851074219,
+      "learning_rate": 4.9625693893735134e-05,
+      "loss": 1.529,
+      "step": 1426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7341501712799072,
+      "learning_rate": 4.962542955326461e-05,
+      "loss": 1.4279,
+      "step": 1427
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3292319774627686,
+      "learning_rate": 4.962516521279408e-05,
+      "loss": 1.5161,
+      "step": 1428
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.153554677963257,
+      "learning_rate": 4.9624900872323555e-05,
+      "loss": 1.5372,
+      "step": 1429
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4493452310562134,
+      "learning_rate": 4.962463653185303e-05,
+      "loss": 1.18,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0727720260620117,
+      "learning_rate": 4.96243721913825e-05,
+      "loss": 1.5504,
+      "step": 1431
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5907639265060425,
+      "learning_rate": 4.9624107850911976e-05,
+      "loss": 1.4157,
+      "step": 1432
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7610145807266235,
+      "learning_rate": 4.9623843510441456e-05,
+      "loss": 1.3546,
+      "step": 1433
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1038999557495117,
+      "learning_rate": 4.962357916997092e-05,
+      "loss": 1.5529,
+      "step": 1434
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5685904026031494,
+      "learning_rate": 4.9623314829500403e-05,
+      "loss": 1.4057,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.773917555809021,
+      "learning_rate": 4.962305048902987e-05,
+      "loss": 1.3452,
+      "step": 1436
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7212398052215576,
+      "learning_rate": 4.9622786148559344e-05,
+      "loss": 1.4391,
+      "step": 1437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7626640796661377,
+      "learning_rate": 4.9622521808088824e-05,
+      "loss": 1.3098,
+      "step": 1438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.054382801055908,
+      "learning_rate": 4.962225746761829e-05,
+      "loss": 1.4525,
+      "step": 1439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6262387037277222,
+      "learning_rate": 4.962199312714777e-05,
+      "loss": 1.4794,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4522370100021362,
+      "learning_rate": 4.962172878667724e-05,
+      "loss": 1.3978,
+      "step": 1441
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4458551406860352,
+      "learning_rate": 4.962146444620672e-05,
+      "loss": 1.4263,
+      "step": 1442
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.409454345703125,
+      "learning_rate": 4.962120010573619e-05,
+      "loss": 1.5138,
+      "step": 1443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6351948976516724,
+      "learning_rate": 4.9620935765265666e-05,
+      "loss": 1.495,
+      "step": 1444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5454893112182617,
+      "learning_rate": 4.962067142479514e-05,
+      "loss": 1.5649,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.066403388977051,
+      "learning_rate": 4.962040708432461e-05,
+      "loss": 1.5173,
+      "step": 1446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3844411373138428,
+      "learning_rate": 4.9620142743854086e-05,
+      "loss": 1.1033,
+      "step": 1447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6488115787506104,
+      "learning_rate": 4.961987840338356e-05,
+      "loss": 1.2811,
+      "step": 1448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.000948190689087,
+      "learning_rate": 4.961961406291303e-05,
+      "loss": 1.794,
+      "step": 1449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.899519443511963,
+      "learning_rate": 4.961934972244251e-05,
+      "loss": 1.8021,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.492762804031372,
+      "learning_rate": 4.961908538197198e-05,
+      "loss": 1.6406,
+      "step": 1451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3945215940475464,
+      "learning_rate": 4.9618821041501454e-05,
+      "loss": 1.3319,
+      "step": 1452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5512380599975586,
+      "learning_rate": 4.9618556701030934e-05,
+      "loss": 1.5284,
+      "step": 1453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7063205242156982,
+      "learning_rate": 4.96182923605604e-05,
+      "loss": 1.4739,
+      "step": 1454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.591920018196106,
+      "learning_rate": 4.961802802008988e-05,
+      "loss": 1.905,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7044870853424072,
+      "learning_rate": 4.961776367961935e-05,
+      "loss": 1.5087,
+      "step": 1456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.94208824634552,
+      "learning_rate": 4.961749933914883e-05,
+      "loss": 1.505,
+      "step": 1457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6600396633148193,
+      "learning_rate": 4.96172349986783e-05,
+      "loss": 1.4358,
+      "step": 1458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5258640050888062,
+      "learning_rate": 4.9616970658207776e-05,
+      "loss": 1.2838,
+      "step": 1459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.292611837387085,
+      "learning_rate": 4.961670631773725e-05,
+      "loss": 1.7064,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6925464868545532,
+      "learning_rate": 4.961644197726672e-05,
+      "loss": 1.4371,
+      "step": 1461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4689338207244873,
+      "learning_rate": 4.9616177636796197e-05,
+      "loss": 1.4797,
+      "step": 1462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4254072904586792,
+      "learning_rate": 4.961591329632567e-05,
+      "loss": 1.323,
+      "step": 1463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4569748640060425,
+      "learning_rate": 4.9615648955855144e-05,
+      "loss": 1.4482,
+      "step": 1464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9684183597564697,
+      "learning_rate": 4.961538461538462e-05,
+      "loss": 1.4627,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7574262619018555,
+      "learning_rate": 4.961512027491409e-05,
+      "loss": 1.3691,
+      "step": 1466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4353163242340088,
+      "learning_rate": 4.9614855934443564e-05,
+      "loss": 1.4099,
+      "step": 1467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7858350276947021,
+      "learning_rate": 4.9614591593973045e-05,
+      "loss": 1.3646,
+      "step": 1468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.021390438079834,
+      "learning_rate": 4.961432725350251e-05,
+      "loss": 1.5903,
+      "step": 1469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7911962270736694,
+      "learning_rate": 4.961406291303199e-05,
+      "loss": 1.5454,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2654597759246826,
+      "learning_rate": 4.961379857256146e-05,
+      "loss": 1.6121,
+      "step": 1471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3775606155395508,
+      "learning_rate": 4.961353423209094e-05,
+      "loss": 1.2901,
+      "step": 1472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9237004518508911,
+      "learning_rate": 4.961326989162041e-05,
+      "loss": 1.5897,
+      "step": 1473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.150212287902832,
+      "learning_rate": 4.961300555114988e-05,
+      "loss": 1.6418,
+      "step": 1474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1737043857574463,
+      "learning_rate": 4.961274121067936e-05,
+      "loss": 1.5692,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5105338096618652,
+      "learning_rate": 4.9612476870208827e-05,
+      "loss": 1.9436,
+      "step": 1476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0150949954986572,
+      "learning_rate": 4.961221252973831e-05,
+      "loss": 1.4348,
+      "step": 1477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7785179615020752,
+      "learning_rate": 4.961194818926778e-05,
+      "loss": 1.4482,
+      "step": 1478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.413366675376892,
+      "learning_rate": 4.9611683848797254e-05,
+      "loss": 1.2517,
+      "step": 1479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5713179111480713,
+      "learning_rate": 4.961141950832673e-05,
+      "loss": 1.4469,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.535345196723938,
+      "learning_rate": 4.96111551678562e-05,
+      "loss": 1.4778,
+      "step": 1481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5654866695404053,
+      "learning_rate": 4.9610890827385675e-05,
+      "loss": 5.403,
+      "step": 1482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.942082166671753,
+      "learning_rate": 4.961062648691515e-05,
+      "loss": 1.6107,
+      "step": 1483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.141674518585205,
+      "learning_rate": 4.961036214644462e-05,
+      "loss": 1.3846,
+      "step": 1484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5543090105056763,
+      "learning_rate": 4.9610097805974095e-05,
+      "loss": 1.4013,
+      "step": 1485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.410356044769287,
+      "learning_rate": 4.960983346550357e-05,
+      "loss": 1.4951,
+      "step": 1486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0547797679901123,
+      "learning_rate": 4.960956912503304e-05,
+      "loss": 1.7685,
+      "step": 1487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.621949315071106,
+      "learning_rate": 4.960930478456252e-05,
+      "loss": 1.4516,
+      "step": 1488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6171997785568237,
+      "learning_rate": 4.960904044409199e-05,
+      "loss": 1.3727,
+      "step": 1489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.687998652458191,
+      "learning_rate": 4.960877610362147e-05,
+      "loss": 1.3585,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.823466420173645,
+      "learning_rate": 4.960851176315094e-05,
+      "loss": 1.4703,
+      "step": 1491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9088199138641357,
+      "learning_rate": 4.960824742268042e-05,
+      "loss": 1.3614,
+      "step": 1492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5815876722335815,
+      "learning_rate": 4.960798308220989e-05,
+      "loss": 1.4492,
+      "step": 1493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8303039073944092,
+      "learning_rate": 4.9607718741739364e-05,
+      "loss": 1.4938,
+      "step": 1494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0490365028381348,
+      "learning_rate": 4.960745440126884e-05,
+      "loss": 1.4221,
+      "step": 1495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9165204763412476,
+      "learning_rate": 4.960719006079831e-05,
+      "loss": 1.5881,
+      "step": 1496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2283014059066772,
+      "learning_rate": 4.9606925720327785e-05,
+      "loss": 1.3371,
+      "step": 1497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.433910608291626,
+      "learning_rate": 4.960666137985726e-05,
+      "loss": 1.4238,
+      "step": 1498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7123323678970337,
+      "learning_rate": 4.960639703938673e-05,
+      "loss": 1.4869,
+      "step": 1499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4784080982208252,
+      "learning_rate": 4.9606132698916206e-05,
+      "loss": 1.3482,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7392635345458984,
+      "learning_rate": 4.960586835844568e-05,
+      "loss": 1.2945,
+      "step": 1501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.509704828262329,
+      "learning_rate": 4.960560401797515e-05,
+      "loss": 1.3892,
+      "step": 1502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4880638122558594,
+      "learning_rate": 4.960533967750463e-05,
+      "loss": 1.4459,
+      "step": 1503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.190537214279175,
+      "learning_rate": 4.96050753370341e-05,
+      "loss": 1.498,
+      "step": 1504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1483938694000244,
+      "learning_rate": 4.960481099656358e-05,
+      "loss": 1.5719,
+      "step": 1505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.650244951248169,
+      "learning_rate": 4.960454665609305e-05,
+      "loss": 1.6542,
+      "step": 1506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9383476972579956,
+      "learning_rate": 4.960428231562253e-05,
+      "loss": 1.5801,
+      "step": 1507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3861944675445557,
+      "learning_rate": 4.9604017975152e-05,
+      "loss": 1.3605,
+      "step": 1508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8627034425735474,
+      "learning_rate": 4.960375363468147e-05,
+      "loss": 1.8025,
+      "step": 1509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7323657274246216,
+      "learning_rate": 4.960348929421095e-05,
+      "loss": 1.5469,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.625321626663208,
+      "learning_rate": 4.9603224953740415e-05,
+      "loss": 1.4939,
+      "step": 1511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6873096227645874,
+      "learning_rate": 4.9602960613269895e-05,
+      "loss": 1.5143,
+      "step": 1512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3846323490142822,
+      "learning_rate": 4.960269627279937e-05,
+      "loss": 1.5457,
+      "step": 1513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6125468015670776,
+      "learning_rate": 4.960243193232884e-05,
+      "loss": 1.4976,
+      "step": 1514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7903023958206177,
+      "learning_rate": 4.9602167591858316e-05,
+      "loss": 2.8314,
+      "step": 1515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.942070484161377,
+      "learning_rate": 4.960190325138779e-05,
+      "loss": 1.4998,
+      "step": 1516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.553556203842163,
+      "learning_rate": 4.960163891091726e-05,
+      "loss": 1.2551,
+      "step": 1517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2704750299453735,
+      "learning_rate": 4.9601374570446744e-05,
+      "loss": 1.3802,
+      "step": 1518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5667221546173096,
+      "learning_rate": 4.960111022997621e-05,
+      "loss": 1.3296,
+      "step": 1519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3207045793533325,
+      "learning_rate": 4.9600845889505684e-05,
+      "loss": 1.3919,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.302562952041626,
+      "learning_rate": 4.960058154903516e-05,
+      "loss": 1.4758,
+      "step": 1521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8370649814605713,
+      "learning_rate": 4.960031720856463e-05,
+      "loss": 1.4732,
+      "step": 1522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.418968915939331,
+      "learning_rate": 4.960005286809411e-05,
+      "loss": 1.3914,
+      "step": 1523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7984591722488403,
+      "learning_rate": 4.959978852762358e-05,
+      "loss": 1.3424,
+      "step": 1524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5587078332901,
+      "learning_rate": 4.959952418715306e-05,
+      "loss": 1.411,
+      "step": 1525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.668395757675171,
+      "learning_rate": 4.9599259846682525e-05,
+      "loss": 1.4765,
+      "step": 1526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7652499675750732,
+      "learning_rate": 4.9598995506212006e-05,
+      "loss": 1.4423,
+      "step": 1527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6073501110076904,
+      "learning_rate": 4.959873116574148e-05,
+      "loss": 1.6017,
+      "step": 1528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.320108652114868,
+      "learning_rate": 4.959846682527095e-05,
+      "loss": 1.5613,
+      "step": 1529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0458102226257324,
+      "learning_rate": 4.9598202484800427e-05,
+      "loss": 1.5654,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5000224113464355,
+      "learning_rate": 4.95979381443299e-05,
+      "loss": 1.239,
+      "step": 1531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4105324745178223,
+      "learning_rate": 4.9597673803859374e-05,
+      "loss": 2.0587,
+      "step": 1532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7553046941757202,
+      "learning_rate": 4.959740946338885e-05,
+      "loss": 1.5451,
+      "step": 1533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5881390571594238,
+      "learning_rate": 4.959714512291832e-05,
+      "loss": 1.366,
+      "step": 1534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4300211668014526,
+      "learning_rate": 4.9596880782447794e-05,
+      "loss": 1.2982,
+      "step": 1535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2717812061309814,
+      "learning_rate": 4.959661644197727e-05,
+      "loss": 1.4183,
+      "step": 1536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8419585227966309,
+      "learning_rate": 4.959635210150674e-05,
+      "loss": 1.4743,
+      "step": 1537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.925929546356201,
+      "learning_rate": 4.959608776103622e-05,
+      "loss": 1.535,
+      "step": 1538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6360324621200562,
+      "learning_rate": 4.959582342056569e-05,
+      "loss": 1.336,
+      "step": 1539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3192601203918457,
+      "learning_rate": 4.959555908009517e-05,
+      "loss": 1.2678,
+      "step": 1540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8304877281188965,
+      "learning_rate": 4.9595294739624636e-05,
+      "loss": 1.3775,
+      "step": 1541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.459349274635315,
+      "learning_rate": 4.9595030399154116e-05,
+      "loss": 1.319,
+      "step": 1542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.144235849380493,
+      "learning_rate": 4.959476605868359e-05,
+      "loss": 1.3691,
+      "step": 1543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7562063932418823,
+      "learning_rate": 4.959450171821306e-05,
+      "loss": 1.4851,
+      "step": 1544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7474582195281982,
+      "learning_rate": 4.959423737774254e-05,
+      "loss": 1.5076,
+      "step": 1545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7668026685714722,
+      "learning_rate": 4.9593973037272004e-05,
+      "loss": 1.4399,
+      "step": 1546
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6016823053359985,
+      "learning_rate": 4.9593708696801484e-05,
+      "loss": 1.3615,
+      "step": 1547
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9639430046081543,
+      "learning_rate": 4.959344435633096e-05,
+      "loss": 1.4052,
+      "step": 1548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7890775203704834,
+      "learning_rate": 4.959318001586043e-05,
+      "loss": 1.4046,
+      "step": 1549
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.501227855682373,
+      "learning_rate": 4.9592915675389905e-05,
+      "loss": 1.8799,
+      "step": 1550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6566779613494873,
+      "learning_rate": 4.959265133491938e-05,
+      "loss": 1.5056,
+      "step": 1551
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2954347133636475,
+      "learning_rate": 4.959238699444885e-05,
+      "loss": 1.6226,
+      "step": 1552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8625856637954712,
+      "learning_rate": 4.959212265397833e-05,
+      "loss": 1.4618,
+      "step": 1553
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.584789514541626,
+      "learning_rate": 4.95918583135078e-05,
+      "loss": 1.4147,
+      "step": 1554
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6849989891052246,
+      "learning_rate": 4.959159397303727e-05,
+      "loss": 1.4816,
+      "step": 1555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7143239974975586,
+      "learning_rate": 4.9591329632566746e-05,
+      "loss": 1.4368,
+      "step": 1556
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4022098779678345,
+      "learning_rate": 4.959106529209622e-05,
+      "loss": 1.3467,
+      "step": 1557
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8715945482254028,
+      "learning_rate": 4.95908009516257e-05,
+      "loss": 1.3514,
+      "step": 1558
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6503853797912598,
+      "learning_rate": 4.959053661115517e-05,
+      "loss": 1.2132,
+      "step": 1559
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.652044296264648,
+      "learning_rate": 4.959027227068465e-05,
+      "loss": 1.4983,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6178275346755981,
+      "learning_rate": 4.9590007930214114e-05,
+      "loss": 1.3743,
+      "step": 1561
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5786150693893433,
+      "learning_rate": 4.9589743589743594e-05,
+      "loss": 1.3808,
+      "step": 1562
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2720184326171875,
+      "learning_rate": 4.958947924927307e-05,
+      "loss": 2.2682,
+      "step": 1563
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.66763436794281,
+      "learning_rate": 4.958921490880254e-05,
+      "loss": 1.6432,
+      "step": 1564
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.381856918334961,
+      "learning_rate": 4.9588950568332015e-05,
+      "loss": 1.6156,
+      "step": 1565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5703303813934326,
+      "learning_rate": 4.958868622786149e-05,
+      "loss": 1.4231,
+      "step": 1566
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4593781232833862,
+      "learning_rate": 4.958842188739096e-05,
+      "loss": 2.4449,
+      "step": 1567
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1835744380950928,
+      "learning_rate": 4.9588157546920436e-05,
+      "loss": 1.3741,
+      "step": 1568
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9764572381973267,
+      "learning_rate": 4.958789320644991e-05,
+      "loss": 1.5541,
+      "step": 1569
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.682018756866455,
+      "learning_rate": 4.958762886597938e-05,
+      "loss": 1.4353,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4604620933532715,
+      "learning_rate": 4.9587364525508856e-05,
+      "loss": 1.2906,
+      "step": 1571
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9639320373535156,
+      "learning_rate": 4.958710018503833e-05,
+      "loss": 1.5322,
+      "step": 1572
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.231764078140259,
+      "learning_rate": 4.958683584456781e-05,
+      "loss": 1.5841,
+      "step": 1573
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4063231945037842,
+      "learning_rate": 4.958657150409728e-05,
+      "loss": 1.4252,
+      "step": 1574
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4082958698272705,
+      "learning_rate": 4.958630716362676e-05,
+      "loss": 1.3434,
+      "step": 1575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3398388624191284,
+      "learning_rate": 4.9586042823156224e-05,
+      "loss": 1.53,
+      "step": 1576
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5701082944869995,
+      "learning_rate": 4.9585778482685705e-05,
+      "loss": 1.4199,
+      "step": 1577
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3495862483978271,
+      "learning_rate": 4.958551414221518e-05,
+      "loss": 1.426,
+      "step": 1578
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.294845700263977,
+      "learning_rate": 4.958524980174465e-05,
+      "loss": 1.2582,
+      "step": 1579
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5897797346115112,
+      "learning_rate": 4.9584985461274125e-05,
+      "loss": 1.4659,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9925825595855713,
+      "learning_rate": 4.958472112080359e-05,
+      "loss": 1.4579,
+      "step": 1581
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2927275896072388,
+      "learning_rate": 4.958445678033307e-05,
+      "loss": 1.4015,
+      "step": 1582
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9567683935165405,
+      "learning_rate": 4.9584192439862546e-05,
+      "loss": 1.3776,
+      "step": 1583
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.169185161590576,
+      "learning_rate": 4.958392809939202e-05,
+      "loss": 1.5591,
+      "step": 1584
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5081685781478882,
+      "learning_rate": 4.958366375892149e-05,
+      "loss": 1.4276,
+      "step": 1585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3420804738998413,
+      "learning_rate": 4.958339941845097e-05,
+      "loss": 2.3592,
+      "step": 1586
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5321258306503296,
+      "learning_rate": 4.958313507798044e-05,
+      "loss": 1.3842,
+      "step": 1587
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.729621410369873,
+      "learning_rate": 4.958287073750992e-05,
+      "loss": 1.2575,
+      "step": 1588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3906311988830566,
+      "learning_rate": 4.958260639703939e-05,
+      "loss": 1.3782,
+      "step": 1589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8676477670669556,
+      "learning_rate": 4.958234205656887e-05,
+      "loss": 1.6034,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3891196250915527,
+      "learning_rate": 4.9582077716098335e-05,
+      "loss": 1.3847,
+      "step": 1591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2640490531921387,
+      "learning_rate": 4.958181337562781e-05,
+      "loss": 1.4102,
+      "step": 1592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5393112897872925,
+      "learning_rate": 4.958154903515729e-05,
+      "loss": 1.4716,
+      "step": 1593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2802183628082275,
+      "learning_rate": 4.9581284694686755e-05,
+      "loss": 1.3371,
+      "step": 1594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7162057161331177,
+      "learning_rate": 4.9581020354216236e-05,
+      "loss": 1.5779,
+      "step": 1595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.148259162902832,
+      "learning_rate": 4.95807560137457e-05,
+      "loss": 1.4671,
+      "step": 1596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.464602470397949,
+      "learning_rate": 4.958049167327518e-05,
+      "loss": 1.6264,
+      "step": 1597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.727612018585205,
+      "learning_rate": 4.9580227332804656e-05,
+      "loss": 1.515,
+      "step": 1598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4732924699783325,
+      "learning_rate": 4.957996299233413e-05,
+      "loss": 1.5484,
+      "step": 1599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2027180194854736,
+      "learning_rate": 4.9579698651863604e-05,
+      "loss": 1.6505,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5612642765045166,
+      "learning_rate": 4.957943431139308e-05,
+      "loss": 1.3786,
+      "step": 1601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9757940769195557,
+      "learning_rate": 4.957916997092255e-05,
+      "loss": 1.4767,
+      "step": 1602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.565129280090332,
+      "learning_rate": 4.9578905630452024e-05,
+      "loss": 1.4529,
+      "step": 1603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6427428722381592,
+      "learning_rate": 4.95786412899815e-05,
+      "loss": 1.549,
+      "step": 1604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5773718357086182,
+      "learning_rate": 4.957837694951097e-05,
+      "loss": 1.6625,
+      "step": 1605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6526451110839844,
+      "learning_rate": 4.9578112609040445e-05,
+      "loss": 1.5916,
+      "step": 1606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0668141841888428,
+      "learning_rate": 4.957784826856992e-05,
+      "loss": 1.6493,
+      "step": 1607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.094925880432129,
+      "learning_rate": 4.95775839280994e-05,
+      "loss": 1.5243,
+      "step": 1608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.706193447113037,
+      "learning_rate": 4.9577319587628866e-05,
+      "loss": 1.5137,
+      "step": 1609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5821620225906372,
+      "learning_rate": 4.9577055247158346e-05,
+      "loss": 1.4054,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8823164701461792,
+      "learning_rate": 4.957679090668781e-05,
+      "loss": 1.7063,
+      "step": 1611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4768058061599731,
+      "learning_rate": 4.957652656621729e-05,
+      "loss": 1.4302,
+      "step": 1612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.89289128780365,
+      "learning_rate": 4.957626222574677e-05,
+      "loss": 1.3009,
+      "step": 1613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6972522735595703,
+      "learning_rate": 4.957599788527624e-05,
+      "loss": 1.6412,
+      "step": 1614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8341267108917236,
+      "learning_rate": 4.9575733544805714e-05,
+      "loss": 1.3644,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4357317686080933,
+      "learning_rate": 4.957546920433519e-05,
+      "loss": 1.3431,
+      "step": 1616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3982058763504028,
+      "learning_rate": 4.957520486386466e-05,
+      "loss": 1.4466,
+      "step": 1617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6967241764068604,
+      "learning_rate": 4.9574940523394135e-05,
+      "loss": 1.4842,
+      "step": 1618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8799158334732056,
+      "learning_rate": 4.957467618292361e-05,
+      "loss": 1.4418,
+      "step": 1619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7508772611618042,
+      "learning_rate": 4.957441184245308e-05,
+      "loss": 2.3858,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.628426194190979,
+      "learning_rate": 4.9574147501982555e-05,
+      "loss": 1.4064,
+      "step": 1621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7519879341125488,
+      "learning_rate": 4.957388316151203e-05,
+      "loss": 1.4196,
+      "step": 1622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9579204320907593,
+      "learning_rate": 4.957361882104151e-05,
+      "loss": 2.0496,
+      "step": 1623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3949480056762695,
+      "learning_rate": 4.9573354480570976e-05,
+      "loss": 1.4873,
+      "step": 1624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6621040105819702,
+      "learning_rate": 4.9573090140100456e-05,
+      "loss": 1.4231,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.173163652420044,
+      "learning_rate": 4.957282579962992e-05,
+      "loss": 1.6363,
+      "step": 1626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5768439769744873,
+      "learning_rate": 4.95725614591594e-05,
+      "loss": 1.3831,
+      "step": 1627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5228557586669922,
+      "learning_rate": 4.957229711868888e-05,
+      "loss": 1.4322,
+      "step": 1628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5580599308013916,
+      "learning_rate": 4.9572032778218344e-05,
+      "loss": 1.3439,
+      "step": 1629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.823197364807129,
+      "learning_rate": 4.9571768437747824e-05,
+      "loss": 1.5573,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4735908508300781,
+      "learning_rate": 4.957150409727729e-05,
+      "loss": 1.4438,
+      "step": 1631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6562364101409912,
+      "learning_rate": 4.957123975680677e-05,
+      "loss": 1.4916,
+      "step": 1632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2361860275268555,
+      "learning_rate": 4.9570975416336245e-05,
+      "loss": 2.2626,
+      "step": 1633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.498873233795166,
+      "learning_rate": 4.957071107586572e-05,
+      "loss": 1.3407,
+      "step": 1634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8398882150650024,
+      "learning_rate": 4.957044673539519e-05,
+      "loss": 1.4868,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.350348949432373,
+      "learning_rate": 4.9570182394924666e-05,
+      "loss": 1.6751,
+      "step": 1636
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6213070154190063,
+      "learning_rate": 4.956991805445414e-05,
+      "loss": 1.2377,
+      "step": 1637
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7495533227920532,
+      "learning_rate": 4.956965371398361e-05,
+      "loss": 1.4521,
+      "step": 1638
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.605144739151001,
+      "learning_rate": 4.9569389373513086e-05,
+      "loss": 1.3463,
+      "step": 1639
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7606133222579956,
+      "learning_rate": 4.956912503304256e-05,
+      "loss": 1.4315,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4992769956588745,
+      "learning_rate": 4.9568860692572034e-05,
+      "loss": 1.3926,
+      "step": 1641
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7759568691253662,
+      "learning_rate": 4.956859635210151e-05,
+      "loss": 1.6169,
+      "step": 1642
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.013232707977295,
+      "learning_rate": 4.956833201163099e-05,
+      "loss": 1.4517,
+      "step": 1643
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8048884868621826,
+      "learning_rate": 4.9568067671160454e-05,
+      "loss": 1.3592,
+      "step": 1644
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7865666151046753,
+      "learning_rate": 4.9567803330689935e-05,
+      "loss": 1.3658,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6202212572097778,
+      "learning_rate": 4.95675389902194e-05,
+      "loss": 1.4322,
+      "step": 1646
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7455825805664062,
+      "learning_rate": 4.956727464974888e-05,
+      "loss": 1.397,
+      "step": 1647
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7802808284759521,
+      "learning_rate": 4.9567010309278355e-05,
+      "loss": 1.3781,
+      "step": 1648
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6419296264648438,
+      "learning_rate": 4.956674596880783e-05,
+      "loss": 1.5422,
+      "step": 1649
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8913413286209106,
+      "learning_rate": 4.95664816283373e-05,
+      "loss": 1.5021,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4184647798538208,
+      "learning_rate": 4.9566217287866776e-05,
+      "loss": 1.3017,
+      "step": 1651
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6011207103729248,
+      "learning_rate": 4.956595294739625e-05,
+      "loss": 1.3876,
+      "step": 1652
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5452293157577515,
+      "learning_rate": 4.956568860692572e-05,
+      "loss": 1.3583,
+      "step": 1653
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7268949747085571,
+      "learning_rate": 4.95654242664552e-05,
+      "loss": 1.5186,
+      "step": 1654
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.706526517868042,
+      "learning_rate": 4.956515992598467e-05,
+      "loss": 1.2949,
+      "step": 1655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.457634449005127,
+      "learning_rate": 4.9564895585514144e-05,
+      "loss": 1.3915,
+      "step": 1656
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6069999933242798,
+      "learning_rate": 4.956463124504362e-05,
+      "loss": 1.2647,
+      "step": 1657
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3370022773742676,
+      "learning_rate": 4.95643669045731e-05,
+      "loss": 1.4192,
+      "step": 1658
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6530468463897705,
+      "learning_rate": 4.9564102564102565e-05,
+      "loss": 1.4813,
+      "step": 1659
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.72675621509552,
+      "learning_rate": 4.9563838223632045e-05,
+      "loss": 1.397,
+      "step": 1660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.345613718032837,
+      "learning_rate": 4.956357388316151e-05,
+      "loss": 1.3981,
+      "step": 1661
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6660879850387573,
+      "learning_rate": 4.956330954269099e-05,
+      "loss": 1.5926,
+      "step": 1662
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7471565008163452,
+      "learning_rate": 4.9563045202220466e-05,
+      "loss": 1.4213,
+      "step": 1663
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6378240585327148,
+      "learning_rate": 4.956278086174993e-05,
+      "loss": 1.4469,
+      "step": 1664
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4509754180908203,
+      "learning_rate": 4.956251652127941e-05,
+      "loss": 1.7389,
+      "step": 1665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2805514335632324,
+      "learning_rate": 4.956225218080888e-05,
+      "loss": 1.348,
+      "step": 1666
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2102258205413818,
+      "learning_rate": 4.956198784033836e-05,
+      "loss": 1.3146,
+      "step": 1667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6394617557525635,
+      "learning_rate": 4.9561723499867834e-05,
+      "loss": 1.4069,
+      "step": 1668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.023862361907959,
+      "learning_rate": 4.956145915939731e-05,
+      "loss": 1.5509,
+      "step": 1669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0046849250793457,
+      "learning_rate": 4.956119481892678e-05,
+      "loss": 1.5209,
+      "step": 1670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6232130527496338,
+      "learning_rate": 4.9560930478456254e-05,
+      "loss": 1.4737,
+      "step": 1671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2027409076690674,
+      "learning_rate": 4.956066613798573e-05,
+      "loss": 1.4301,
+      "step": 1672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4766788482666016,
+      "learning_rate": 4.95604017975152e-05,
+      "loss": 1.4295,
+      "step": 1673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.285869598388672,
+      "learning_rate": 4.9560137457044675e-05,
+      "loss": 1.5557,
+      "step": 1674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4509947299957275,
+      "learning_rate": 4.955987311657415e-05,
+      "loss": 2.817,
+      "step": 1675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6586003303527832,
+      "learning_rate": 4.955960877610362e-05,
+      "loss": 1.4454,
+      "step": 1676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6795002222061157,
+      "learning_rate": 4.9559344435633096e-05,
+      "loss": 1.4813,
+      "step": 1677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4418399333953857,
+      "learning_rate": 4.9559080095162576e-05,
+      "loss": 1.5429,
+      "step": 1678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4673309326171875,
+      "learning_rate": 4.955881575469204e-05,
+      "loss": 1.4068,
+      "step": 1679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.625831127166748,
+      "learning_rate": 4.955855141422152e-05,
+      "loss": 1.3238,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1055102348327637,
+      "learning_rate": 4.955828707375099e-05,
+      "loss": 1.3949,
+      "step": 1681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8717643022537231,
+      "learning_rate": 4.955802273328047e-05,
+      "loss": 1.4648,
+      "step": 1682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3826571702957153,
+      "learning_rate": 4.9557758392809944e-05,
+      "loss": 1.3759,
+      "step": 1683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2895822525024414,
+      "learning_rate": 4.955749405233942e-05,
+      "loss": 1.4655,
+      "step": 1684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8304845094680786,
+      "learning_rate": 4.955722971186889e-05,
+      "loss": 1.3502,
+      "step": 1685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9586799144744873,
+      "learning_rate": 4.9556965371398365e-05,
+      "loss": 1.2552,
+      "step": 1686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3766279220581055,
+      "learning_rate": 4.955670103092784e-05,
+      "loss": 1.883,
+      "step": 1687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9413634538650513,
+      "learning_rate": 4.955643669045731e-05,
+      "loss": 1.5308,
+      "step": 1688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5934746265411377,
+      "learning_rate": 4.9556172349986785e-05,
+      "loss": 1.478,
+      "step": 1689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1187126636505127,
+      "learning_rate": 4.955590800951626e-05,
+      "loss": 1.4976,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.83664071559906,
+      "learning_rate": 4.955564366904573e-05,
+      "loss": 1.5558,
+      "step": 1691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7629085779190063,
+      "learning_rate": 4.9555379328575206e-05,
+      "loss": 1.3861,
+      "step": 1692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9662811756134033,
+      "learning_rate": 4.9555114988104686e-05,
+      "loss": 1.3695,
+      "step": 1693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7148569822311401,
+      "learning_rate": 4.955485064763415e-05,
+      "loss": 1.3988,
+      "step": 1694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0393054485321045,
+      "learning_rate": 4.9554586307163633e-05,
+      "loss": 1.4765,
+      "step": 1695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3136650323867798,
+      "learning_rate": 4.95543219666931e-05,
+      "loss": 1.3258,
+      "step": 1696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.552286386489868,
+      "learning_rate": 4.955405762622258e-05,
+      "loss": 1.3324,
+      "step": 1697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4684550762176514,
+      "learning_rate": 4.9553793285752054e-05,
+      "loss": 1.3924,
+      "step": 1698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8305892944335938,
+      "learning_rate": 4.955352894528152e-05,
+      "loss": 1.3544,
+      "step": 1699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6300088167190552,
+      "learning_rate": 4.9553264604811e-05,
+      "loss": 1.3973,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5596847534179688,
+      "learning_rate": 4.955300026434047e-05,
+      "loss": 1.4692,
+      "step": 1701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8836950063705444,
+      "learning_rate": 4.955273592386995e-05,
+      "loss": 1.453,
+      "step": 1702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2650684118270874,
+      "learning_rate": 4.955247158339942e-05,
+      "loss": 1.2015,
+      "step": 1703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3971197605133057,
+      "learning_rate": 4.9552207242928896e-05,
+      "loss": 1.4827,
+      "step": 1704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8300282955169678,
+      "learning_rate": 4.955194290245837e-05,
+      "loss": 1.3003,
+      "step": 1705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2823666334152222,
+      "learning_rate": 4.955167856198784e-05,
+      "loss": 1.4362,
+      "step": 1706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5497201681137085,
+      "learning_rate": 4.9551414221517316e-05,
+      "loss": 1.5454,
+      "step": 1707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5173053741455078,
+      "learning_rate": 4.95511498810468e-05,
+      "loss": 1.3636,
+      "step": 1708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7285562753677368,
+      "learning_rate": 4.9550885540576263e-05,
+      "loss": 1.4134,
+      "step": 1709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5004671812057495,
+      "learning_rate": 4.955062120010574e-05,
+      "loss": 1.3412,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.291914939880371,
+      "learning_rate": 4.955035685963521e-05,
+      "loss": 1.8089,
+      "step": 1711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4912961721420288,
+      "learning_rate": 4.9550092519164684e-05,
+      "loss": 1.3897,
+      "step": 1712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5509449243545532,
+      "learning_rate": 4.9549828178694165e-05,
+      "loss": 1.4935,
+      "step": 1713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8999574184417725,
+      "learning_rate": 4.954956383822363e-05,
+      "loss": 1.5513,
+      "step": 1714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5513269901275635,
+      "learning_rate": 4.954929949775311e-05,
+      "loss": 1.4011,
+      "step": 1715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5226551294326782,
+      "learning_rate": 4.954903515728258e-05,
+      "loss": 1.4177,
+      "step": 1716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0362331867218018,
+      "learning_rate": 4.954877081681206e-05,
+      "loss": 1.514,
+      "step": 1717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3188315629959106,
+      "learning_rate": 4.954850647634153e-05,
+      "loss": 1.2244,
+      "step": 1718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7171486616134644,
+      "learning_rate": 4.9548242135871006e-05,
+      "loss": 1.4219,
+      "step": 1719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6470202207565308,
+      "learning_rate": 4.954797779540048e-05,
+      "loss": 1.425,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2368409633636475,
+      "learning_rate": 4.954771345492995e-05,
+      "loss": 1.5334,
+      "step": 1721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5998183488845825,
+      "learning_rate": 4.954744911445943e-05,
+      "loss": 1.3292,
+      "step": 1722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8922221660614014,
+      "learning_rate": 4.95471847739889e-05,
+      "loss": 1.3763,
+      "step": 1723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7813023328781128,
+      "learning_rate": 4.9546920433518374e-05,
+      "loss": 1.4308,
+      "step": 1724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.674144983291626,
+      "learning_rate": 4.954665609304785e-05,
+      "loss": 1.4306,
+      "step": 1725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.005150079727173,
+      "learning_rate": 4.954639175257732e-05,
+      "loss": 1.341,
+      "step": 1726
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6207860708236694,
+      "learning_rate": 4.9546127412106795e-05,
+      "loss": 2.4876,
+      "step": 1727
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6342620849609375,
+      "learning_rate": 4.9545863071636275e-05,
+      "loss": 1.3951,
+      "step": 1728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.787139892578125,
+      "learning_rate": 4.954559873116574e-05,
+      "loss": 1.4709,
+      "step": 1729
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7001159191131592,
+      "learning_rate": 4.954533439069522e-05,
+      "loss": 1.4008,
+      "step": 1730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0887200832366943,
+      "learning_rate": 4.954507005022469e-05,
+      "loss": 1.5799,
+      "step": 1731
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2182955741882324,
+      "learning_rate": 4.954480570975417e-05,
+      "loss": 1.7978,
+      "step": 1732
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.572948694229126,
+      "learning_rate": 4.954454136928364e-05,
+      "loss": 1.4927,
+      "step": 1733
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4186747074127197,
+      "learning_rate": 4.9544277028813116e-05,
+      "loss": 1.2798,
+      "step": 1734
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7752355337142944,
+      "learning_rate": 4.954401268834259e-05,
+      "loss": 1.4502,
+      "step": 1735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8524038791656494,
+      "learning_rate": 4.954374834787206e-05,
+      "loss": 1.4703,
+      "step": 1736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.000239372253418,
+      "learning_rate": 4.954348400740154e-05,
+      "loss": 1.3188,
+      "step": 1737
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.538757562637329,
+      "learning_rate": 4.954321966693101e-05,
+      "loss": 1.5356,
+      "step": 1738
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.657509207725525,
+      "learning_rate": 4.9542955326460484e-05,
+      "loss": 1.4845,
+      "step": 1739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.154041051864624,
+      "learning_rate": 4.954269098598996e-05,
+      "loss": 2.6838,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3936692476272583,
+      "learning_rate": 4.954242664551943e-05,
+      "loss": 1.2398,
+      "step": 1741
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.77518630027771,
+      "learning_rate": 4.9542162305048905e-05,
+      "loss": 1.3177,
+      "step": 1742
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2939972877502441,
+      "learning_rate": 4.9541897964578385e-05,
+      "loss": 1.4679,
+      "step": 1743
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7911241054534912,
+      "learning_rate": 4.954163362410785e-05,
+      "loss": 1.327,
+      "step": 1744
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6598881483078003,
+      "learning_rate": 4.9541369283637326e-05,
+      "loss": 1.3879,
+      "step": 1745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4625364542007446,
+      "learning_rate": 4.95411049431668e-05,
+      "loss": 1.2949,
+      "step": 1746
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.366725444793701,
+      "learning_rate": 4.954084060269627e-05,
+      "loss": 1.6545,
+      "step": 1747
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.649012804031372,
+      "learning_rate": 4.954057626222575e-05,
+      "loss": 1.3954,
+      "step": 1748
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4067076444625854,
+      "learning_rate": 4.954031192175522e-05,
+      "loss": 1.412,
+      "step": 1749
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.83164381980896,
+      "learning_rate": 4.95400475812847e-05,
+      "loss": 1.3336,
+      "step": 1750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7586134672164917,
+      "learning_rate": 4.953978324081417e-05,
+      "loss": 1.4312,
+      "step": 1751
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.376784801483154,
+      "learning_rate": 4.953951890034365e-05,
+      "loss": 1.6352,
+      "step": 1752
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0272765159606934,
+      "learning_rate": 4.953925455987312e-05,
+      "loss": 1.29,
+      "step": 1753
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8198583126068115,
+      "learning_rate": 4.9538990219402594e-05,
+      "loss": 1.4767,
+      "step": 1754
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.350748062133789,
+      "learning_rate": 4.953872587893207e-05,
+      "loss": 1.5046,
+      "step": 1755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4138638973236084,
+      "learning_rate": 4.953846153846154e-05,
+      "loss": 1.5038,
+      "step": 1756
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.767543911933899,
+      "learning_rate": 4.9538197197991015e-05,
+      "loss": 1.4631,
+      "step": 1757
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9695860147476196,
+      "learning_rate": 4.953793285752049e-05,
+      "loss": 1.3029,
+      "step": 1758
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9194743633270264,
+      "learning_rate": 4.953766851704996e-05,
+      "loss": 1.4529,
+      "step": 1759
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0775387287139893,
+      "learning_rate": 4.9537404176579436e-05,
+      "loss": 1.6087,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6344115734100342,
+      "learning_rate": 4.953713983610891e-05,
+      "loss": 1.3808,
+      "step": 1761
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4166120290756226,
+      "learning_rate": 4.953687549563838e-05,
+      "loss": 1.3829,
+      "step": 1762
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3311381340026855,
+      "learning_rate": 4.9536611155167863e-05,
+      "loss": 1.1801,
+      "step": 1763
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8351454734802246,
+      "learning_rate": 4.953634681469733e-05,
+      "loss": 1.4226,
+      "step": 1764
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5231029987335205,
+      "learning_rate": 4.953608247422681e-05,
+      "loss": 1.3389,
+      "step": 1765
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5737935304641724,
+      "learning_rate": 4.953581813375628e-05,
+      "loss": 1.4076,
+      "step": 1766
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1866092681884766,
+      "learning_rate": 4.953555379328576e-05,
+      "loss": 1.1717,
+      "step": 1767
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5830085277557373,
+      "learning_rate": 4.953528945281523e-05,
+      "loss": 1.4491,
+      "step": 1768
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5236769914627075,
+      "learning_rate": 4.9535025112344705e-05,
+      "loss": 1.3684,
+      "step": 1769
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6971954107284546,
+      "learning_rate": 4.953476077187418e-05,
+      "loss": 1.3857,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.479791522026062,
+      "learning_rate": 4.953449643140365e-05,
+      "loss": 1.4996,
+      "step": 1771
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8232975006103516,
+      "learning_rate": 4.9534232090933126e-05,
+      "loss": 1.3839,
+      "step": 1772
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7472007274627686,
+      "learning_rate": 4.95339677504626e-05,
+      "loss": 1.4091,
+      "step": 1773
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4389468431472778,
+      "learning_rate": 4.953370340999207e-05,
+      "loss": 1.2926,
+      "step": 1774
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7678273916244507,
+      "learning_rate": 4.9533439069521546e-05,
+      "loss": 1.3508,
+      "step": 1775
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.593605875968933,
+      "learning_rate": 4.953317472905102e-05,
+      "loss": 1.4776,
+      "step": 1776
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5497421026229858,
+      "learning_rate": 4.9532910388580493e-05,
+      "loss": 2.436,
+      "step": 1777
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4951884746551514,
+      "learning_rate": 4.9532646048109974e-05,
+      "loss": 1.2846,
+      "step": 1778
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6524747610092163,
+      "learning_rate": 4.953238170763944e-05,
+      "loss": 1.3773,
+      "step": 1779
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8119590282440186,
+      "learning_rate": 4.953211736716892e-05,
+      "loss": 2.313,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8384443521499634,
+      "learning_rate": 4.953185302669839e-05,
+      "loss": 1.3968,
+      "step": 1781
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.625750184059143,
+      "learning_rate": 4.953158868622786e-05,
+      "loss": 1.4129,
+      "step": 1782
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4617078304290771,
+      "learning_rate": 4.953132434575734e-05,
+      "loss": 1.3846,
+      "step": 1783
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9800794124603271,
+      "learning_rate": 4.953106000528681e-05,
+      "loss": 1.3641,
+      "step": 1784
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6570056676864624,
+      "learning_rate": 4.953079566481629e-05,
+      "loss": 1.4437,
+      "step": 1785
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8459309339523315,
+      "learning_rate": 4.9530531324345756e-05,
+      "loss": 1.5503,
+      "step": 1786
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6540088653564453,
+      "learning_rate": 4.9530266983875236e-05,
+      "loss": 1.395,
+      "step": 1787
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.656670331954956,
+      "learning_rate": 4.95300026434047e-05,
+      "loss": 1.4564,
+      "step": 1788
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8177740573883057,
+      "learning_rate": 4.952973830293418e-05,
+      "loss": 1.3166,
+      "step": 1789
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3160853385925293,
+      "learning_rate": 4.9529473962463657e-05,
+      "loss": 1.5321,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6499154567718506,
+      "learning_rate": 4.952920962199313e-05,
+      "loss": 1.3476,
+      "step": 1791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3169560432434082,
+      "learning_rate": 4.9528945281522604e-05,
+      "loss": 1.4484,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.542883276939392,
+      "learning_rate": 4.952868094105208e-05,
+      "loss": 1.5187,
+      "step": 1793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3903756141662598,
+      "learning_rate": 4.952841660058155e-05,
+      "loss": 1.2665,
+      "step": 1794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6874139308929443,
+      "learning_rate": 4.9528152260111024e-05,
+      "loss": 1.4061,
+      "step": 1795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1368579864501953,
+      "learning_rate": 4.95278879196405e-05,
+      "loss": 1.6162,
+      "step": 1796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5094330310821533,
+      "learning_rate": 4.952762357916997e-05,
+      "loss": 1.4926,
+      "step": 1797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4672267436981201,
+      "learning_rate": 4.9527359238699445e-05,
+      "loss": 1.3558,
+      "step": 1798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.631202220916748,
+      "learning_rate": 4.952709489822892e-05,
+      "loss": 1.5113,
+      "step": 1799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.316175699234009,
+      "learning_rate": 4.95268305577584e-05,
+      "loss": 1.5943,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2846295833587646,
+      "learning_rate": 4.9526566217287866e-05,
+      "loss": 1.6791,
+      "step": 1801
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2980411052703857,
+      "learning_rate": 4.9526301876817346e-05,
+      "loss": 1.3885,
+      "step": 1802
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4460365772247314,
+      "learning_rate": 4.952603753634681e-05,
+      "loss": 1.3807,
+      "step": 1803
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5812824964523315,
+      "learning_rate": 4.952577319587629e-05,
+      "loss": 1.4329,
+      "step": 1804
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.507485032081604,
+      "learning_rate": 4.952550885540577e-05,
+      "loss": 1.6407,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3151718378067017,
+      "learning_rate": 4.952524451493524e-05,
+      "loss": 1.2323,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9206887483596802,
+      "learning_rate": 4.9524980174464714e-05,
+      "loss": 1.6777,
+      "step": 1807
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3340935707092285,
+      "learning_rate": 4.952471583399418e-05,
+      "loss": 2.0382,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.530827522277832,
+      "learning_rate": 4.952445149352366e-05,
+      "loss": 1.3855,
+      "step": 1809
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6346089839935303,
+      "learning_rate": 4.9524187153053135e-05,
+      "loss": 1.4034,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5986751317977905,
+      "learning_rate": 4.952392281258261e-05,
+      "loss": 1.2719,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.384442090988159,
+      "learning_rate": 4.952365847211208e-05,
+      "loss": 1.5718,
+      "step": 1812
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5976186990737915,
+      "learning_rate": 4.9523394131641555e-05,
+      "loss": 1.4628,
+      "step": 1813
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8616909980773926,
+      "learning_rate": 4.952312979117103e-05,
+      "loss": 1.3517,
+      "step": 1814
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.792817234992981,
+      "learning_rate": 4.952286545070051e-05,
+      "loss": 1.5218,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3488694429397583,
+      "learning_rate": 4.9522601110229976e-05,
+      "loss": 1.3882,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2931201457977295,
+      "learning_rate": 4.952233676975945e-05,
+      "loss": 1.5151,
+      "step": 1817
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.843849778175354,
+      "learning_rate": 4.952207242928892e-05,
+      "loss": 1.4169,
+      "step": 1818
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.972848653793335,
+      "learning_rate": 4.95218080888184e-05,
+      "loss": 1.5919,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.636197805404663,
+      "learning_rate": 4.952154374834788e-05,
+      "loss": 1.4663,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6678545475006104,
+      "learning_rate": 4.9521279407877344e-05,
+      "loss": 1.3993,
+      "step": 1821
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5104093551635742,
+      "learning_rate": 4.9521015067406824e-05,
+      "loss": 1.3025,
+      "step": 1822
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8574490547180176,
+      "learning_rate": 4.952075072693629e-05,
+      "loss": 1.5141,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7124594449996948,
+      "learning_rate": 4.952048638646577e-05,
+      "loss": 1.3749,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3348126411437988,
+      "learning_rate": 4.9520222045995245e-05,
+      "loss": 1.3186,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5110782384872437,
+      "learning_rate": 4.951995770552472e-05,
+      "loss": 2.4184,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5042001008987427,
+      "learning_rate": 4.951969336505419e-05,
+      "loss": 1.3336,
+      "step": 1827
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4295719861984253,
+      "learning_rate": 4.9519429024583666e-05,
+      "loss": 1.589,
+      "step": 1828
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5208604335784912,
+      "learning_rate": 4.951916468411314e-05,
+      "loss": 1.5159,
+      "step": 1829
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9461307525634766,
+      "learning_rate": 4.951890034364261e-05,
+      "loss": 1.2895,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9015246629714966,
+      "learning_rate": 4.9518636003172087e-05,
+      "loss": 1.3733,
+      "step": 1831
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0689470767974854,
+      "learning_rate": 4.951837166270156e-05,
+      "loss": 2.5674,
+      "step": 1832
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0428497791290283,
+      "learning_rate": 4.9518107322231034e-05,
+      "loss": 1.7652,
+      "step": 1833
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0728540420532227,
+      "learning_rate": 4.951784298176051e-05,
+      "loss": 1.5716,
+      "step": 1834
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5101308822631836,
+      "learning_rate": 4.951757864128999e-05,
+      "loss": 1.3318,
+      "step": 1835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7114837169647217,
+      "learning_rate": 4.9517314300819454e-05,
+      "loss": 1.4973,
+      "step": 1836
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6239460706710815,
+      "learning_rate": 4.9517049960348935e-05,
+      "loss": 1.5248,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.336622476577759,
+      "learning_rate": 4.95167856198784e-05,
+      "loss": 1.5,
+      "step": 1838
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2913081645965576,
+      "learning_rate": 4.951652127940788e-05,
+      "loss": 1.2225,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3971893787384033,
+      "learning_rate": 4.9516256938937355e-05,
+      "loss": 1.3817,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4877114295959473,
+      "learning_rate": 4.951599259846683e-05,
+      "loss": 1.3801,
+      "step": 1841
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4158321619033813,
+      "learning_rate": 4.95157282579963e-05,
+      "loss": 1.3389,
+      "step": 1842
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.549731969833374,
+      "learning_rate": 4.9515463917525776e-05,
+      "loss": 1.4226,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9393571615219116,
+      "learning_rate": 4.951519957705525e-05,
+      "loss": 1.4471,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9300258159637451,
+      "learning_rate": 4.951493523658472e-05,
+      "loss": 1.6083,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.387586236000061,
+      "learning_rate": 4.95146708961142e-05,
+      "loss": 1.4036,
+      "step": 1846
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4687211513519287,
+      "learning_rate": 4.951440655564367e-05,
+      "loss": 1.1131,
+      "step": 1847
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9156323671340942,
+      "learning_rate": 4.9514142215173144e-05,
+      "loss": 1.4164,
+      "step": 1848
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5583438873291016,
+      "learning_rate": 4.951387787470262e-05,
+      "loss": 1.4122,
+      "step": 1849
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.082967519760132,
+      "learning_rate": 4.95136135342321e-05,
+      "loss": 1.4617,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8827747106552124,
+      "learning_rate": 4.9513349193761565e-05,
+      "loss": 1.5395,
+      "step": 1851
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3311502933502197,
+      "learning_rate": 4.9513084853291045e-05,
+      "loss": 1.4987,
+      "step": 1852
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7298641204833984,
+      "learning_rate": 4.951282051282051e-05,
+      "loss": 1.7,
+      "step": 1853
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8355867862701416,
+      "learning_rate": 4.9512556172349985e-05,
+      "loss": 1.5909,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1845898628234863,
+      "learning_rate": 4.9512291831879466e-05,
+      "loss": 1.6278,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.875148892402649,
+      "learning_rate": 4.951202749140893e-05,
+      "loss": 1.4936,
+      "step": 1856
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.115445375442505,
+      "learning_rate": 4.951176315093841e-05,
+      "loss": 1.5772,
+      "step": 1857
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9172861576080322,
+      "learning_rate": 4.951149881046788e-05,
+      "loss": 1.5321,
+      "step": 1858
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5402171611785889,
+      "learning_rate": 4.951123446999736e-05,
+      "loss": 1.3419,
+      "step": 1859
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.584789752960205,
+      "learning_rate": 4.9510970129526834e-05,
+      "loss": 1.4173,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6987345218658447,
+      "learning_rate": 4.951070578905631e-05,
+      "loss": 1.3869,
+      "step": 1861
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3889784812927246,
+      "learning_rate": 4.951044144858578e-05,
+      "loss": 1.2967,
+      "step": 1862
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0929205417633057,
+      "learning_rate": 4.9510177108115254e-05,
+      "loss": 1.4045,
+      "step": 1863
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5896776914596558,
+      "learning_rate": 4.950991276764473e-05,
+      "loss": 1.3103,
+      "step": 1864
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2326607704162598,
+      "learning_rate": 4.95096484271742e-05,
+      "loss": 1.4512,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3352446556091309,
+      "learning_rate": 4.9509384086703675e-05,
+      "loss": 1.4053,
+      "step": 1866
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.471527099609375,
+      "learning_rate": 4.950911974623315e-05,
+      "loss": 1.5416,
+      "step": 1867
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7600709199905396,
+      "learning_rate": 4.950885540576262e-05,
+      "loss": 1.5965,
+      "step": 1868
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6875473260879517,
+      "learning_rate": 4.9508591065292096e-05,
+      "loss": 1.503,
+      "step": 1869
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.288785457611084,
+      "learning_rate": 4.9508326724821576e-05,
+      "loss": 1.4818,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7068334817886353,
+      "learning_rate": 4.950806238435104e-05,
+      "loss": 1.3233,
+      "step": 1871
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6695400476455688,
+      "learning_rate": 4.950779804388052e-05,
+      "loss": 1.3755,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8470958471298218,
+      "learning_rate": 4.950753370340999e-05,
+      "loss": 1.5816,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7197637557983398,
+      "learning_rate": 4.950726936293947e-05,
+      "loss": 1.3581,
+      "step": 1874
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8507122993469238,
+      "learning_rate": 4.9507005022468944e-05,
+      "loss": 1.3528,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8414294719696045,
+      "learning_rate": 4.950674068199842e-05,
+      "loss": 1.4204,
+      "step": 1876
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1364150047302246,
+      "learning_rate": 4.950647634152789e-05,
+      "loss": 1.4437,
+      "step": 1877
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2107406854629517,
+      "learning_rate": 4.9506212001057365e-05,
+      "loss": 1.2535,
+      "step": 1878
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.829789161682129,
+      "learning_rate": 4.950594766058684e-05,
+      "loss": 1.5743,
+      "step": 1879
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9080663919448853,
+      "learning_rate": 4.950568332011631e-05,
+      "loss": 1.7168,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8417246341705322,
+      "learning_rate": 4.9505418979645785e-05,
+      "loss": 1.584,
+      "step": 1881
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.333091378211975,
+      "learning_rate": 4.950515463917526e-05,
+      "loss": 1.4457,
+      "step": 1882
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.512927770614624,
+      "learning_rate": 4.950489029870473e-05,
+      "loss": 1.3252,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6003384590148926,
+      "learning_rate": 4.9504625958234206e-05,
+      "loss": 1.4738,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8690884113311768,
+      "learning_rate": 4.9504361617763686e-05,
+      "loss": 1.4612,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5306353569030762,
+      "learning_rate": 4.950409727729315e-05,
+      "loss": 1.3746,
+      "step": 1886
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.333751678466797,
+      "learning_rate": 4.9503832936822634e-05,
+      "loss": 1.6141,
+      "step": 1887
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.893277883529663,
+      "learning_rate": 4.95035685963521e-05,
+      "loss": 1.5552,
+      "step": 1888
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.695858955383301,
+      "learning_rate": 4.950330425588158e-05,
+      "loss": 1.3395,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7158764600753784,
+      "learning_rate": 4.9503039915411054e-05,
+      "loss": 1.4717,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1240737438201904,
+      "learning_rate": 4.950277557494052e-05,
+      "loss": 1.4865,
+      "step": 1891
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5714703798294067,
+      "learning_rate": 4.950251123447e-05,
+      "loss": 1.4256,
+      "step": 1892
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8708761930465698,
+      "learning_rate": 4.950224689399947e-05,
+      "loss": 1.5288,
+      "step": 1893
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5926486253738403,
+      "learning_rate": 4.950198255352895e-05,
+      "loss": 1.3099,
+      "step": 1894
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3334203958511353,
+      "learning_rate": 4.950171821305842e-05,
+      "loss": 1.3744,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3688631057739258,
+      "learning_rate": 4.9501453872587896e-05,
+      "loss": 1.3617,
+      "step": 1896
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.415266752243042,
+      "learning_rate": 4.950118953211737e-05,
+      "loss": 1.4839,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3636882305145264,
+      "learning_rate": 4.950092519164684e-05,
+      "loss": 1.5639,
+      "step": 1898
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4615070819854736,
+      "learning_rate": 4.9500660851176316e-05,
+      "loss": 1.2483,
+      "step": 1899
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6492843627929688,
+      "learning_rate": 4.950039651070579e-05,
+      "loss": 1.2899,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8592655658721924,
+      "learning_rate": 4.9500132170235264e-05,
+      "loss": 1.9259,
+      "step": 1901
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5525996685028076,
+      "learning_rate": 4.949986782976474e-05,
+      "loss": 1.7005,
+      "step": 1902
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4929763078689575,
+      "learning_rate": 4.949960348929421e-05,
+      "loss": 1.4329,
+      "step": 1903
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7152099609375,
+      "learning_rate": 4.9499339148823684e-05,
+      "loss": 1.4698,
+      "step": 1904
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6704425811767578,
+      "learning_rate": 4.9499074808353165e-05,
+      "loss": 1.3327,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5394225120544434,
+      "learning_rate": 4.949881046788263e-05,
+      "loss": 1.3327,
+      "step": 1906
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0335516929626465,
+      "learning_rate": 4.949854612741211e-05,
+      "loss": 1.5164,
+      "step": 1907
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.030790090560913,
+      "learning_rate": 4.949828178694158e-05,
+      "loss": 1.6515,
+      "step": 1908
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.928964138031006,
+      "learning_rate": 4.949801744647106e-05,
+      "loss": 1.6447,
+      "step": 1909
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7481236457824707,
+      "learning_rate": 4.949775310600053e-05,
+      "loss": 1.5172,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9886245727539062,
+      "learning_rate": 4.9497488765530006e-05,
+      "loss": 1.3682,
+      "step": 1911
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6402148008346558,
+      "learning_rate": 4.949722442505948e-05,
+      "loss": 1.5014,
+      "step": 1912
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.629797101020813,
+      "learning_rate": 4.949696008458895e-05,
+      "loss": 1.4246,
+      "step": 1913
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.260244607925415,
+      "learning_rate": 4.949669574411843e-05,
+      "loss": 1.4463,
+      "step": 1914
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.561496376991272,
+      "learning_rate": 4.94964314036479e-05,
+      "loss": 1.3755,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7817354202270508,
+      "learning_rate": 4.9496167063177374e-05,
+      "loss": 1.58,
+      "step": 1916
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.640842080116272,
+      "learning_rate": 4.949590272270685e-05,
+      "loss": 1.4849,
+      "step": 1917
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3865160942077637,
+      "learning_rate": 4.949563838223632e-05,
+      "loss": 1.3856,
+      "step": 1918
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5249301195144653,
+      "learning_rate": 4.9495374041765795e-05,
+      "loss": 1.2846,
+      "step": 1919
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.945265293121338,
+      "learning_rate": 4.9495109701295275e-05,
+      "loss": 1.4084,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.723296880722046,
+      "learning_rate": 4.949484536082474e-05,
+      "loss": 1.5706,
+      "step": 1921
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.161287546157837,
+      "learning_rate": 4.949458102035422e-05,
+      "loss": 1.5335,
+      "step": 1922
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.509121060371399,
+      "learning_rate": 4.949431667988369e-05,
+      "loss": 1.4371,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.402935266494751,
+      "learning_rate": 4.949405233941317e-05,
+      "loss": 1.3406,
+      "step": 1924
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.467971920967102,
+      "learning_rate": 4.949378799894264e-05,
+      "loss": 1.4722,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5086190700531006,
+      "learning_rate": 4.949352365847211e-05,
+      "loss": 1.3905,
+      "step": 1926
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1997973918914795,
+      "learning_rate": 4.949325931800159e-05,
+      "loss": 1.6293,
+      "step": 1927
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0799176692962646,
+      "learning_rate": 4.949299497753106e-05,
+      "loss": 1.639,
+      "step": 1928
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.028113842010498,
+      "learning_rate": 4.949273063706054e-05,
+      "loss": 1.5999,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5672273635864258,
+      "learning_rate": 4.949246629659001e-05,
+      "loss": 1.4175,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5024306774139404,
+      "learning_rate": 4.9492201956119484e-05,
+      "loss": 1.4422,
+      "step": 1931
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8687714338302612,
+      "learning_rate": 4.949193761564896e-05,
+      "loss": 1.3924,
+      "step": 1932
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7434871196746826,
+      "learning_rate": 4.949167327517843e-05,
+      "loss": 1.4175,
+      "step": 1933
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6563923358917236,
+      "learning_rate": 4.9491408934707905e-05,
+      "loss": 1.5245,
+      "step": 1934
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9522745609283447,
+      "learning_rate": 4.949114459423738e-05,
+      "loss": 1.3667,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.729227066040039,
+      "learning_rate": 4.949088025376685e-05,
+      "loss": 1.5069,
+      "step": 1936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5696872472763062,
+      "learning_rate": 4.9490615913296326e-05,
+      "loss": 1.363,
+      "step": 1937
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4332557916641235,
+      "learning_rate": 4.94903515728258e-05,
+      "loss": 1.3447,
+      "step": 1938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5612276792526245,
+      "learning_rate": 4.949008723235527e-05,
+      "loss": 1.4208,
+      "step": 1939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5190232992172241,
+      "learning_rate": 4.948982289188475e-05,
+      "loss": 1.4719,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.405017375946045,
+      "learning_rate": 4.948955855141422e-05,
+      "loss": 1.6366,
+      "step": 1941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8180983066558838,
+      "learning_rate": 4.94892942109437e-05,
+      "loss": 1.6598,
+      "step": 1942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6863248348236084,
+      "learning_rate": 4.948902987047317e-05,
+      "loss": 1.3631,
+      "step": 1943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6124482154846191,
+      "learning_rate": 4.948876553000265e-05,
+      "loss": 1.2959,
+      "step": 1944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6836512088775635,
+      "learning_rate": 4.948850118953212e-05,
+      "loss": 1.2536,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4001473188400269,
+      "learning_rate": 4.9488236849061595e-05,
+      "loss": 1.2634,
+      "step": 1946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5169042348861694,
+      "learning_rate": 4.948797250859107e-05,
+      "loss": 1.9111,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6686288118362427,
+      "learning_rate": 4.948770816812054e-05,
+      "loss": 1.3653,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6784685850143433,
+      "learning_rate": 4.9487443827650015e-05,
+      "loss": 1.3916,
+      "step": 1949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3487303256988525,
+      "learning_rate": 4.948717948717949e-05,
+      "loss": 1.7845,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.46384596824646,
+      "learning_rate": 4.948691514670896e-05,
+      "loss": 1.3056,
+      "step": 1951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7411103248596191,
+      "learning_rate": 4.9486650806238436e-05,
+      "loss": 1.635,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4644980430603027,
+      "learning_rate": 4.948638646576791e-05,
+      "loss": 1.6977,
+      "step": 1953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.471846342086792,
+      "learning_rate": 4.948612212529738e-05,
+      "loss": 1.2561,
+      "step": 1954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.458645224571228,
+      "learning_rate": 4.9485857784826864e-05,
+      "loss": 1.1338,
+      "step": 1955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.468119740486145,
+      "learning_rate": 4.948559344435633e-05,
+      "loss": 1.3505,
+      "step": 1956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7300660610198975,
+      "learning_rate": 4.948532910388581e-05,
+      "loss": 1.3979,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4192390441894531,
+      "learning_rate": 4.948506476341528e-05,
+      "loss": 1.3671,
+      "step": 1958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.662264108657837,
+      "learning_rate": 4.948480042294476e-05,
+      "loss": 1.5366,
+      "step": 1959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6586499214172363,
+      "learning_rate": 4.948453608247423e-05,
+      "loss": 1.576,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9684011936187744,
+      "learning_rate": 4.9484271742003705e-05,
+      "loss": 1.484,
+      "step": 1961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7603631019592285,
+      "learning_rate": 4.948400740153318e-05,
+      "loss": 1.1024,
+      "step": 1962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.855921745300293,
+      "learning_rate": 4.9483743061062645e-05,
+      "loss": 1.4652,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9098448753356934,
+      "learning_rate": 4.9483478720592126e-05,
+      "loss": 1.6708,
+      "step": 1964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0738725662231445,
+      "learning_rate": 4.94832143801216e-05,
+      "loss": 1.5524,
+      "step": 1965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6745588779449463,
+      "learning_rate": 4.948295003965107e-05,
+      "loss": 1.2975,
+      "step": 1966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7258206605911255,
+      "learning_rate": 4.9482685699180546e-05,
+      "loss": 1.3079,
+      "step": 1967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8991971015930176,
+      "learning_rate": 4.948242135871002e-05,
+      "loss": 1.4277,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.455250859260559,
+      "learning_rate": 4.9482157018239494e-05,
+      "loss": 1.377,
+      "step": 1969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1902843713760376,
+      "learning_rate": 4.9481892677768974e-05,
+      "loss": 1.218,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.468750476837158,
+      "learning_rate": 4.948162833729844e-05,
+      "loss": 1.4162,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6346486806869507,
+      "learning_rate": 4.9481363996827914e-05,
+      "loss": 1.8332,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.511103630065918,
+      "learning_rate": 4.948109965635739e-05,
+      "loss": 1.3482,
+      "step": 1973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8295890092849731,
+      "learning_rate": 4.948083531588686e-05,
+      "loss": 1.5408,
+      "step": 1974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3582921028137207,
+      "learning_rate": 4.948057097541634e-05,
+      "loss": 1.2314,
+      "step": 1975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8337591886520386,
+      "learning_rate": 4.948030663494581e-05,
+      "loss": 1.4975,
+      "step": 1976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7335623502731323,
+      "learning_rate": 4.948004229447529e-05,
+      "loss": 1.4879,
+      "step": 1977
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7746957540512085,
+      "learning_rate": 4.9479777954004756e-05,
+      "loss": 1.31,
+      "step": 1978
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8422696590423584,
+      "learning_rate": 4.9479513613534236e-05,
+      "loss": 1.5529,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2964868545532227,
+      "learning_rate": 4.947924927306371e-05,
+      "loss": 1.47,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8993827104568481,
+      "learning_rate": 4.947898493259318e-05,
+      "loss": 1.6053,
+      "step": 1981
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9730879068374634,
+      "learning_rate": 4.947872059212266e-05,
+      "loss": 1.594,
+      "step": 1982
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.245389461517334,
+      "learning_rate": 4.947845625165213e-05,
+      "loss": 1.6124,
+      "step": 1983
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8077224493026733,
+      "learning_rate": 4.9478191911181604e-05,
+      "loss": 1.6056,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.774619460105896,
+      "learning_rate": 4.947792757071108e-05,
+      "loss": 1.4967,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.712030291557312,
+      "learning_rate": 4.947766323024055e-05,
+      "loss": 1.3405,
+      "step": 1986
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.499290943145752,
+      "learning_rate": 4.9477398889770025e-05,
+      "loss": 1.4844,
+      "step": 1987
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6018627882003784,
+      "learning_rate": 4.94771345492995e-05,
+      "loss": 1.418,
+      "step": 1988
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5035669803619385,
+      "learning_rate": 4.947687020882897e-05,
+      "loss": 1.5349,
+      "step": 1989
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6551947593688965,
+      "learning_rate": 4.947660586835845e-05,
+      "loss": 1.4426,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.424117922782898,
+      "learning_rate": 4.947634152788792e-05,
+      "loss": 1.4031,
+      "step": 1991
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4983242750167847,
+      "learning_rate": 4.94760771874174e-05,
+      "loss": 1.4046,
+      "step": 1992
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.24371075630188,
+      "learning_rate": 4.9475812846946866e-05,
+      "loss": 1.4046,
+      "step": 1993
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1068077087402344,
+      "learning_rate": 4.9475548506476346e-05,
+      "loss": 1.5223,
+      "step": 1994
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.039043426513672,
+      "learning_rate": 4.947528416600582e-05,
+      "loss": 1.5195,
+      "step": 1995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.070765972137451,
+      "learning_rate": 4.9475019825535294e-05,
+      "loss": 1.3208,
+      "step": 1996
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7839536666870117,
+      "learning_rate": 4.947475548506477e-05,
+      "loss": 1.8334,
+      "step": 1997
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6909633874893188,
+      "learning_rate": 4.9474491144594234e-05,
+      "loss": 1.3796,
+      "step": 1998
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6867072582244873,
+      "learning_rate": 4.9474226804123714e-05,
+      "loss": 1.3098,
+      "step": 1999
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8114961385726929,
+      "learning_rate": 4.947396246365319e-05,
+      "loss": 1.5963,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0715668201446533,
+      "learning_rate": 4.947369812318266e-05,
+      "loss": 1.3835,
+      "step": 2001
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.605961799621582,
+      "learning_rate": 4.9473433782712135e-05,
+      "loss": 1.3492,
+      "step": 2002
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.022421360015869,
+      "learning_rate": 4.947316944224161e-05,
+      "loss": 1.4512,
+      "step": 2003
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3593881130218506,
+      "learning_rate": 4.947290510177108e-05,
+      "loss": 1.467,
+      "step": 2004
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9061893224716187,
+      "learning_rate": 4.947264076130056e-05,
+      "loss": 1.3395,
+      "step": 2005
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8164234161376953,
+      "learning_rate": 4.947237642083003e-05,
+      "loss": 1.5766,
+      "step": 2006
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.60530686378479,
+      "learning_rate": 4.94721120803595e-05,
+      "loss": 1.3849,
+      "step": 2007
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.355069875717163,
+      "learning_rate": 4.9471847739888976e-05,
+      "loss": 1.3587,
+      "step": 2008
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4219653606414795,
+      "learning_rate": 4.947158339941845e-05,
+      "loss": 1.3416,
+      "step": 2009
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7992854118347168,
+      "learning_rate": 4.947131905894793e-05,
+      "loss": 1.4946,
+      "step": 2010
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9473320245742798,
+      "learning_rate": 4.94710547184774e-05,
+      "loss": 1.213,
+      "step": 2011
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3872570991516113,
+      "learning_rate": 4.947079037800688e-05,
+      "loss": 1.6437,
+      "step": 2012
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.68748140335083,
+      "learning_rate": 4.9470526037536344e-05,
+      "loss": 2.2569,
+      "step": 2013
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6049507856369019,
+      "learning_rate": 4.9470261697065825e-05,
+      "loss": 1.324,
+      "step": 2014
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.375252366065979,
+      "learning_rate": 4.94699973565953e-05,
+      "loss": 1.2926,
+      "step": 2015
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4445960521697998,
+      "learning_rate": 4.946973301612477e-05,
+      "loss": 1.2879,
+      "step": 2016
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0743188858032227,
+      "learning_rate": 4.9469468675654245e-05,
+      "loss": 1.6131,
+      "step": 2017
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5538479089736938,
+      "learning_rate": 4.946920433518372e-05,
+      "loss": 1.4177,
+      "step": 2018
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7586251497268677,
+      "learning_rate": 4.946893999471319e-05,
+      "loss": 2.7339,
+      "step": 2019
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8020941019058228,
+      "learning_rate": 4.9468675654242666e-05,
+      "loss": 1.4063,
+      "step": 2020
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.684766173362732,
+      "learning_rate": 4.946841131377214e-05,
+      "loss": 1.4356,
+      "step": 2021
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8827513456344604,
+      "learning_rate": 4.946814697330161e-05,
+      "loss": 1.3963,
+      "step": 2022
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5985850095748901,
+      "learning_rate": 4.946788263283109e-05,
+      "loss": 1.477,
+      "step": 2023
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.318706750869751,
+      "learning_rate": 4.946761829236056e-05,
+      "loss": 1.6249,
+      "step": 2024
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.892790675163269,
+      "learning_rate": 4.946735395189004e-05,
+      "loss": 1.4491,
+      "step": 2025
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5069752931594849,
+      "learning_rate": 4.946708961141951e-05,
+      "loss": 1.3485,
+      "step": 2026
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7036848068237305,
+      "learning_rate": 4.946682527094899e-05,
+      "loss": 1.3985,
+      "step": 2027
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.568444848060608,
+      "learning_rate": 4.9466560930478455e-05,
+      "loss": 1.3294,
+      "step": 2028
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4605220556259155,
+      "learning_rate": 4.9466296590007935e-05,
+      "loss": 1.5791,
+      "step": 2029
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5584118366241455,
+      "learning_rate": 4.946603224953741e-05,
+      "loss": 1.5575,
+      "step": 2030
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.875349044799805,
+      "learning_rate": 4.946576790906688e-05,
+      "loss": 1.5363,
+      "step": 2031
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.291548013687134,
+      "learning_rate": 4.9465503568596356e-05,
+      "loss": 1.5813,
+      "step": 2032
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.864281177520752,
+      "learning_rate": 4.946523922812583e-05,
+      "loss": 1.468,
+      "step": 2033
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4756131172180176,
+      "learning_rate": 4.94649748876553e-05,
+      "loss": 1.3101,
+      "step": 2034
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9442154169082642,
+      "learning_rate": 4.9464710547184776e-05,
+      "loss": 1.5101,
+      "step": 2035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2426055669784546,
+      "learning_rate": 4.946444620671425e-05,
+      "loss": 1.3674,
+      "step": 2036
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7085434198379517,
+      "learning_rate": 4.9464181866243723e-05,
+      "loss": 1.5111,
+      "step": 2037
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.338965892791748,
+      "learning_rate": 4.94639175257732e-05,
+      "loss": 1.6443,
+      "step": 2038
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6230461597442627,
+      "learning_rate": 4.946365318530267e-05,
+      "loss": 1.5082,
+      "step": 2039
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5564851760864258,
+      "learning_rate": 4.946338884483215e-05,
+      "loss": 1.3819,
+      "step": 2040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6455879211425781,
+      "learning_rate": 4.946312450436162e-05,
+      "loss": 1.4023,
+      "step": 2041
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5488477945327759,
+      "learning_rate": 4.94628601638911e-05,
+      "loss": 1.4179,
+      "step": 2042
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6850264072418213,
+      "learning_rate": 4.9462595823420565e-05,
+      "loss": 2.2832,
+      "step": 2043
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5671181678771973,
+      "learning_rate": 4.946233148295004e-05,
+      "loss": 1.3327,
+      "step": 2044
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.832033157348633,
+      "learning_rate": 4.946206714247952e-05,
+      "loss": 1.4343,
+      "step": 2045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7656123638153076,
+      "learning_rate": 4.9461802802008986e-05,
+      "loss": 1.429,
+      "step": 2046
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.531612515449524,
+      "learning_rate": 4.9461538461538466e-05,
+      "loss": 1.3643,
+      "step": 2047
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5404921770095825,
+      "learning_rate": 4.946127412106793e-05,
+      "loss": 1.387,
+      "step": 2048
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6397268772125244,
+      "learning_rate": 4.946100978059741e-05,
+      "loss": 1.6191,
+      "step": 2049
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.363858699798584,
+      "learning_rate": 4.946074544012689e-05,
+      "loss": 1.4534,
+      "step": 2050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.465752124786377,
+      "learning_rate": 4.946048109965636e-05,
+      "loss": 1.2609,
+      "step": 2051
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7237741947174072,
+      "learning_rate": 4.9460216759185834e-05,
+      "loss": 1.4478,
+      "step": 2052
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7529200315475464,
+      "learning_rate": 4.945995241871531e-05,
+      "loss": 1.4119,
+      "step": 2053
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6335200071334839,
+      "learning_rate": 4.945968807824478e-05,
+      "loss": 1.343,
+      "step": 2054
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5942423343658447,
+      "learning_rate": 4.9459423737774255e-05,
+      "loss": 1.4854,
+      "step": 2055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.293573021888733,
+      "learning_rate": 4.945915939730373e-05,
+      "loss": 1.2749,
+      "step": 2056
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5511082410812378,
+      "learning_rate": 4.94588950568332e-05,
+      "loss": 1.3051,
+      "step": 2057
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6929445266723633,
+      "learning_rate": 4.9458630716362675e-05,
+      "loss": 1.3536,
+      "step": 2058
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6030625104904175,
+      "learning_rate": 4.945836637589215e-05,
+      "loss": 1.4539,
+      "step": 2059
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2590208053588867,
+      "learning_rate": 4.945810203542163e-05,
+      "loss": 1.5268,
+      "step": 2060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7412580251693726,
+      "learning_rate": 4.9457837694951096e-05,
+      "loss": 1.7263,
+      "step": 2061
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4557276964187622,
+      "learning_rate": 4.9457573354480576e-05,
+      "loss": 1.367,
+      "step": 2062
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6003447771072388,
+      "learning_rate": 4.945730901401004e-05,
+      "loss": 1.4351,
+      "step": 2063
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7181293964385986,
+      "learning_rate": 4.9457044673539523e-05,
+      "loss": 1.3548,
+      "step": 2064
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.681015133857727,
+      "learning_rate": 4.9456780333069e-05,
+      "loss": 1.3855,
+      "step": 2065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.923283338546753,
+      "learning_rate": 4.945651599259847e-05,
+      "loss": 1.464,
+      "step": 2066
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.090954542160034,
+      "learning_rate": 4.9456251652127944e-05,
+      "loss": 1.514,
+      "step": 2067
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7853879928588867,
+      "learning_rate": 4.945598731165742e-05,
+      "loss": 1.6095,
+      "step": 2068
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.447334051132202,
+      "learning_rate": 4.945572297118689e-05,
+      "loss": 2.3753,
+      "step": 2069
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7427910566329956,
+      "learning_rate": 4.9455458630716365e-05,
+      "loss": 1.4874,
+      "step": 2070
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8327977657318115,
+      "learning_rate": 4.945519429024584e-05,
+      "loss": 1.5258,
+      "step": 2071
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9750118255615234,
+      "learning_rate": 4.945492994977531e-05,
+      "loss": 1.6845,
+      "step": 2072
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.791123390197754,
+      "learning_rate": 4.9454665609304786e-05,
+      "loss": 1.6122,
+      "step": 2073
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8987687826156616,
+      "learning_rate": 4.945440126883426e-05,
+      "loss": 1.3685,
+      "step": 2074
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7751656770706177,
+      "learning_rate": 4.945413692836374e-05,
+      "loss": 1.5158,
+      "step": 2075
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.55458664894104,
+      "learning_rate": 4.9453872587893206e-05,
+      "loss": 1.288,
+      "step": 2076
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3966821432113647,
+      "learning_rate": 4.945360824742269e-05,
+      "loss": 1.5881,
+      "step": 2077
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7340995073318481,
+      "learning_rate": 4.9453343906952153e-05,
+      "loss": 1.4733,
+      "step": 2078
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9633578062057495,
+      "learning_rate": 4.9453079566481634e-05,
+      "loss": 1.5484,
+      "step": 2079
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.671161413192749,
+      "learning_rate": 4.945281522601111e-05,
+      "loss": 1.3213,
+      "step": 2080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5570343732833862,
+      "learning_rate": 4.9452550885540574e-05,
+      "loss": 1.4248,
+      "step": 2081
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9307961463928223,
+      "learning_rate": 4.9452286545070054e-05,
+      "loss": 1.5105,
+      "step": 2082
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.782128930091858,
+      "learning_rate": 4.945202220459952e-05,
+      "loss": 1.5016,
+      "step": 2083
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.370213747024536,
+      "learning_rate": 4.9451757864129e-05,
+      "loss": 1.2904,
+      "step": 2084
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7416105270385742,
+      "learning_rate": 4.9451493523658475e-05,
+      "loss": 1.3788,
+      "step": 2085
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.032334566116333,
+      "learning_rate": 4.945122918318795e-05,
+      "loss": 1.4876,
+      "step": 2086
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.652172565460205,
+      "learning_rate": 4.945096484271742e-05,
+      "loss": 1.4298,
+      "step": 2087
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8130356073379517,
+      "learning_rate": 4.9450700502246896e-05,
+      "loss": 1.6195,
+      "step": 2088
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9606739282608032,
+      "learning_rate": 4.945043616177637e-05,
+      "loss": 1.4585,
+      "step": 2089
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6567435264587402,
+      "learning_rate": 4.945017182130584e-05,
+      "loss": 1.2892,
+      "step": 2090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8529447317123413,
+      "learning_rate": 4.944990748083532e-05,
+      "loss": 1.4487,
+      "step": 2091
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.633589744567871,
+      "learning_rate": 4.944964314036479e-05,
+      "loss": 1.5077,
+      "step": 2092
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9611443281173706,
+      "learning_rate": 4.9449378799894264e-05,
+      "loss": 1.3652,
+      "step": 2093
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8058758974075317,
+      "learning_rate": 4.944911445942374e-05,
+      "loss": 1.3723,
+      "step": 2094
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4812217950820923,
+      "learning_rate": 4.944885011895322e-05,
+      "loss": 1.1457,
+      "step": 2095
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6531181335449219,
+      "learning_rate": 4.9448585778482684e-05,
+      "loss": 1.4206,
+      "step": 2096
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7102638483047485,
+      "learning_rate": 4.9448321438012165e-05,
+      "loss": 1.3418,
+      "step": 2097
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9775288105010986,
+      "learning_rate": 4.944805709754163e-05,
+      "loss": 1.3873,
+      "step": 2098
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0794014930725098,
+      "learning_rate": 4.944779275707111e-05,
+      "loss": 1.4426,
+      "step": 2099
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2029132843017578,
+      "learning_rate": 4.9447528416600586e-05,
+      "loss": 1.263,
+      "step": 2100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.14691162109375,
+      "learning_rate": 4.944726407613006e-05,
+      "loss": 1.5292,
+      "step": 2101
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9558658599853516,
+      "learning_rate": 4.944699973565953e-05,
+      "loss": 1.3617,
+      "step": 2102
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.421854019165039,
+      "learning_rate": 4.9446735395189006e-05,
+      "loss": 1.3803,
+      "step": 2103
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9749516248703003,
+      "learning_rate": 4.944647105471848e-05,
+      "loss": 1.5474,
+      "step": 2104
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9370917081832886,
+      "learning_rate": 4.9446206714247953e-05,
+      "loss": 1.532,
+      "step": 2105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6472392082214355,
+      "learning_rate": 4.944594237377743e-05,
+      "loss": 1.4715,
+      "step": 2106
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6807024478912354,
+      "learning_rate": 4.94456780333069e-05,
+      "loss": 1.4407,
+      "step": 2107
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.229849100112915,
+      "learning_rate": 4.9445413692836374e-05,
+      "loss": 1.8011,
+      "step": 2108
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.099060297012329,
+      "learning_rate": 4.944514935236585e-05,
+      "loss": 1.3555,
+      "step": 2109
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.941178560256958,
+      "learning_rate": 4.944488501189533e-05,
+      "loss": 1.6047,
+      "step": 2110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6210730075836182,
+      "learning_rate": 4.9444620671424795e-05,
+      "loss": 1.5413,
+      "step": 2111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6555845737457275,
+      "learning_rate": 4.9444356330954275e-05,
+      "loss": 1.5777,
+      "step": 2112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6885582208633423,
+      "learning_rate": 4.944409199048374e-05,
+      "loss": 1.446,
+      "step": 2113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9277195930480957,
+      "learning_rate": 4.944382765001322e-05,
+      "loss": 1.3647,
+      "step": 2114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0771663188934326,
+      "learning_rate": 4.9443563309542696e-05,
+      "loss": 1.447,
+      "step": 2115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5923384428024292,
+      "learning_rate": 4.944329896907216e-05,
+      "loss": 1.3898,
+      "step": 2116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.694679617881775,
+      "learning_rate": 4.944303462860164e-05,
+      "loss": 1.3317,
+      "step": 2117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6755801439285278,
+      "learning_rate": 4.944277028813111e-05,
+      "loss": 2.1459,
+      "step": 2118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6459851264953613,
+      "learning_rate": 4.944250594766059e-05,
+      "loss": 1.4583,
+      "step": 2119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.757878065109253,
+      "learning_rate": 4.9442241607190064e-05,
+      "loss": 1.4131,
+      "step": 2120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6704280376434326,
+      "learning_rate": 4.944197726671954e-05,
+      "loss": 1.3809,
+      "step": 2121
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5235735177993774,
+      "learning_rate": 4.944171292624901e-05,
+      "loss": 1.4767,
+      "step": 2122
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5474307537078857,
+      "learning_rate": 4.9441448585778484e-05,
+      "loss": 1.4313,
+      "step": 2123
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7302480936050415,
+      "learning_rate": 4.944118424530796e-05,
+      "loss": 1.4614,
+      "step": 2124
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6585263013839722,
+      "learning_rate": 4.944091990483743e-05,
+      "loss": 1.2784,
+      "step": 2125
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9361553192138672,
+      "learning_rate": 4.9440655564366905e-05,
+      "loss": 1.4122,
+      "step": 2126
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.646461844444275,
+      "learning_rate": 4.944039122389638e-05,
+      "loss": 1.3498,
+      "step": 2127
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8552016019821167,
+      "learning_rate": 4.944012688342585e-05,
+      "loss": 1.901,
+      "step": 2128
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3645823001861572,
+      "learning_rate": 4.9439862542955326e-05,
+      "loss": 1.1544,
+      "step": 2129
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5724250078201294,
+      "learning_rate": 4.9439598202484806e-05,
+      "loss": 1.3478,
+      "step": 2130
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4671133756637573,
+      "learning_rate": 4.943933386201427e-05,
+      "loss": 1.3294,
+      "step": 2131
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.511041283607483,
+      "learning_rate": 4.943906952154375e-05,
+      "loss": 1.4882,
+      "step": 2132
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1203231811523438,
+      "learning_rate": 4.943880518107322e-05,
+      "loss": 1.5079,
+      "step": 2133
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7689199447631836,
+      "learning_rate": 4.94385408406027e-05,
+      "loss": 1.6043,
+      "step": 2134
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7023117542266846,
+      "learning_rate": 4.9438276500132174e-05,
+      "loss": 1.4702,
+      "step": 2135
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6602880954742432,
+      "learning_rate": 4.943801215966165e-05,
+      "loss": 1.5026,
+      "step": 2136
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.254461407661438,
+      "learning_rate": 4.943774781919112e-05,
+      "loss": 1.2525,
+      "step": 2137
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6968204975128174,
+      "learning_rate": 4.9437483478720595e-05,
+      "loss": 1.4062,
+      "step": 2138
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9700262546539307,
+      "learning_rate": 4.943721913825007e-05,
+      "loss": 1.536,
+      "step": 2139
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8937747478485107,
+      "learning_rate": 4.943695479777954e-05,
+      "loss": 1.3105,
+      "step": 2140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.514430284500122,
+      "learning_rate": 4.9436690457309016e-05,
+      "loss": 1.5785,
+      "step": 2141
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5991542339324951,
+      "learning_rate": 4.943642611683849e-05,
+      "loss": 1.3175,
+      "step": 2142
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.010817527770996,
+      "learning_rate": 4.943616177636796e-05,
+      "loss": 1.4612,
+      "step": 2143
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.192291021347046,
+      "learning_rate": 4.9435897435897436e-05,
+      "loss": 1.4024,
+      "step": 2144
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1878371238708496,
+      "learning_rate": 4.9435633095426917e-05,
+      "loss": 1.8266,
+      "step": 2145
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8938915729522705,
+      "learning_rate": 4.943536875495638e-05,
+      "loss": 1.4922,
+      "step": 2146
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.679075002670288,
+      "learning_rate": 4.9435104414485864e-05,
+      "loss": 1.4831,
+      "step": 2147
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4871646165847778,
+      "learning_rate": 4.943484007401533e-05,
+      "loss": 1.2798,
+      "step": 2148
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6656934022903442,
+      "learning_rate": 4.943457573354481e-05,
+      "loss": 1.3552,
+      "step": 2149
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6633509397506714,
+      "learning_rate": 4.9434311393074284e-05,
+      "loss": 1.6154,
+      "step": 2150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8230605125427246,
+      "learning_rate": 4.943404705260376e-05,
+      "loss": 1.4632,
+      "step": 2151
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.614266276359558,
+      "learning_rate": 4.943378271213323e-05,
+      "loss": 1.3207,
+      "step": 2152
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.418208360671997,
+      "learning_rate": 4.94335183716627e-05,
+      "loss": 1.5646,
+      "step": 2153
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.503174066543579,
+      "learning_rate": 4.943325403119218e-05,
+      "loss": 1.3487,
+      "step": 2154
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4023691415786743,
+      "learning_rate": 4.943298969072165e-05,
+      "loss": 1.2753,
+      "step": 2155
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0077733993530273,
+      "learning_rate": 4.9432725350251126e-05,
+      "loss": 1.5203,
+      "step": 2156
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7383995056152344,
+      "learning_rate": 4.94324610097806e-05,
+      "loss": 1.4732,
+      "step": 2157
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9039878845214844,
+      "learning_rate": 4.943219666931007e-05,
+      "loss": 1.4128,
+      "step": 2158
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.662673830986023,
+      "learning_rate": 4.9431932328839547e-05,
+      "loss": 1.333,
+      "step": 2159
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.466816782951355,
+      "learning_rate": 4.943166798836903e-05,
+      "loss": 1.374,
+      "step": 2160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0618391036987305,
+      "learning_rate": 4.9431403647898494e-05,
+      "loss": 1.4622,
+      "step": 2161
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4112905263900757,
+      "learning_rate": 4.943113930742797e-05,
+      "loss": 1.2583,
+      "step": 2162
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3022708892822266,
+      "learning_rate": 4.943087496695744e-05,
+      "loss": 1.4494,
+      "step": 2163
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.116779088973999,
+      "learning_rate": 4.9430610626486914e-05,
+      "loss": 1.5184,
+      "step": 2164
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1000280380249023,
+      "learning_rate": 4.9430346286016395e-05,
+      "loss": 1.5383,
+      "step": 2165
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6528466939926147,
+      "learning_rate": 4.943008194554586e-05,
+      "loss": 1.5365,
+      "step": 2166
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5431371927261353,
+      "learning_rate": 4.942981760507534e-05,
+      "loss": 1.3808,
+      "step": 2167
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7998476028442383,
+      "learning_rate": 4.942955326460481e-05,
+      "loss": 1.5697,
+      "step": 2168
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8847758769989014,
+      "learning_rate": 4.942928892413429e-05,
+      "loss": 1.4985,
+      "step": 2169
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9894094467163086,
+      "learning_rate": 4.942902458366376e-05,
+      "loss": 1.4977,
+      "step": 2170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.671401023864746,
+      "learning_rate": 4.9428760243193236e-05,
+      "loss": 1.5329,
+      "step": 2171
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6515511274337769,
+      "learning_rate": 4.942849590272271e-05,
+      "loss": 1.4883,
+      "step": 2172
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6776823997497559,
+      "learning_rate": 4.942823156225218e-05,
+      "loss": 1.3526,
+      "step": 2173
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9013068675994873,
+      "learning_rate": 4.942796722178166e-05,
+      "loss": 1.5063,
+      "step": 2174
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5527665615081787,
+      "learning_rate": 4.942770288131113e-05,
+      "loss": 1.4161,
+      "step": 2175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7885215282440186,
+      "learning_rate": 4.9427438540840604e-05,
+      "loss": 1.4205,
+      "step": 2176
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.590899109840393,
+      "learning_rate": 4.942717420037008e-05,
+      "loss": 1.4108,
+      "step": 2177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.831560730934143,
+      "learning_rate": 4.942690985989955e-05,
+      "loss": 1.5225,
+      "step": 2178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.156463384628296,
+      "learning_rate": 4.9426645519429025e-05,
+      "loss": 1.5016,
+      "step": 2179
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2360601425170898,
+      "learning_rate": 4.9426381178958505e-05,
+      "loss": 1.2082,
+      "step": 2180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.733704686164856,
+      "learning_rate": 4.942611683848797e-05,
+      "loss": 1.3375,
+      "step": 2181
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5879409313201904,
+      "learning_rate": 4.942585249801745e-05,
+      "loss": 1.5005,
+      "step": 2182
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5776376724243164,
+      "learning_rate": 4.942558815754692e-05,
+      "loss": 1.3585,
+      "step": 2183
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1943140029907227,
+      "learning_rate": 4.94253238170764e-05,
+      "loss": 1.5601,
+      "step": 2184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5760620832443237,
+      "learning_rate": 4.942505947660587e-05,
+      "loss": 1.4304,
+      "step": 2185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5470664501190186,
+      "learning_rate": 4.9424795136135347e-05,
+      "loss": 1.4407,
+      "step": 2186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.868588924407959,
+      "learning_rate": 4.942453079566482e-05,
+      "loss": 1.5162,
+      "step": 2187
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4765346050262451,
+      "learning_rate": 4.942426645519429e-05,
+      "loss": 1.323,
+      "step": 2188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7613537311553955,
+      "learning_rate": 4.942400211472377e-05,
+      "loss": 1.4419,
+      "step": 2189
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6956346035003662,
+      "learning_rate": 4.942373777425324e-05,
+      "loss": 1.5194,
+      "step": 2190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6555765867233276,
+      "learning_rate": 4.9423473433782714e-05,
+      "loss": 1.3986,
+      "step": 2191
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5221011638641357,
+      "learning_rate": 4.942320909331219e-05,
+      "loss": 1.4965,
+      "step": 2192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.531356692314148,
+      "learning_rate": 4.942294475284166e-05,
+      "loss": 1.394,
+      "step": 2193
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.213090419769287,
+      "learning_rate": 4.9422680412371135e-05,
+      "loss": 1.3815,
+      "step": 2194
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.072310447692871,
+      "learning_rate": 4.9422416071900615e-05,
+      "loss": 1.5574,
+      "step": 2195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.977637529373169,
+      "learning_rate": 4.942215173143008e-05,
+      "loss": 1.6443,
+      "step": 2196
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4696118831634521,
+      "learning_rate": 4.942188739095956e-05,
+      "loss": 1.5284,
+      "step": 2197
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7614383697509766,
+      "learning_rate": 4.942162305048903e-05,
+      "loss": 1.4414,
+      "step": 2198
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.528731346130371,
+      "learning_rate": 4.94213587100185e-05,
+      "loss": 1.7357,
+      "step": 2199
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3903335332870483,
+      "learning_rate": 4.942109436954798e-05,
+      "loss": 1.4731,
+      "step": 2200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5264685153961182,
+      "learning_rate": 4.942083002907745e-05,
+      "loss": 1.1787,
+      "step": 2201
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.665756106376648,
+      "learning_rate": 4.942056568860693e-05,
+      "loss": 1.4239,
+      "step": 2202
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3922370672225952,
+      "learning_rate": 4.94203013481364e-05,
+      "loss": 1.463,
+      "step": 2203
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.625560998916626,
+      "learning_rate": 4.942003700766588e-05,
+      "loss": 1.3768,
+      "step": 2204
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4528063535690308,
+      "learning_rate": 4.941977266719535e-05,
+      "loss": 1.3159,
+      "step": 2205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7027206420898438,
+      "learning_rate": 4.9419508326724825e-05,
+      "loss": 1.3673,
+      "step": 2206
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5905812978744507,
+      "learning_rate": 4.94192439862543e-05,
+      "loss": 1.3092,
+      "step": 2207
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0169289112091064,
+      "learning_rate": 4.941897964578377e-05,
+      "loss": 1.4976,
+      "step": 2208
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5845757722854614,
+      "learning_rate": 4.9418715305313245e-05,
+      "loss": 1.9442,
+      "step": 2209
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9886729717254639,
+      "learning_rate": 4.941845096484272e-05,
+      "loss": 1.4244,
+      "step": 2210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3601577281951904,
+      "learning_rate": 4.941818662437219e-05,
+      "loss": 1.9599,
+      "step": 2211
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.512617349624634,
+      "learning_rate": 4.9417922283901666e-05,
+      "loss": 1.5052,
+      "step": 2212
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6150870323181152,
+      "learning_rate": 4.941765794343114e-05,
+      "loss": 1.4198,
+      "step": 2213
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8111244440078735,
+      "learning_rate": 4.941739360296061e-05,
+      "loss": 1.6655,
+      "step": 2214
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5842921733856201,
+      "learning_rate": 4.9417129262490094e-05,
+      "loss": 1.3181,
+      "step": 2215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8328300714492798,
+      "learning_rate": 4.941686492201956e-05,
+      "loss": 1.4573,
+      "step": 2216
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.95075523853302,
+      "learning_rate": 4.941660058154904e-05,
+      "loss": 1.464,
+      "step": 2217
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.217282295227051,
+      "learning_rate": 4.941633624107851e-05,
+      "loss": 1.5625,
+      "step": 2218
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4643396139144897,
+      "learning_rate": 4.941607190060799e-05,
+      "loss": 1.334,
+      "step": 2219
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8565380573272705,
+      "learning_rate": 4.941580756013746e-05,
+      "loss": 1.4167,
+      "step": 2220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9488811492919922,
+      "learning_rate": 4.9415543219666935e-05,
+      "loss": 1.4663,
+      "step": 2221
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6648262739181519,
+      "learning_rate": 4.941527887919641e-05,
+      "loss": 1.3282,
+      "step": 2222
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.543783187866211,
+      "learning_rate": 4.941501453872588e-05,
+      "loss": 1.5008,
+      "step": 2223
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4663865566253662,
+      "learning_rate": 4.9414750198255356e-05,
+      "loss": 1.316,
+      "step": 2224
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6117268800735474,
+      "learning_rate": 4.941448585778483e-05,
+      "loss": 1.3699,
+      "step": 2225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1691691875457764,
+      "learning_rate": 4.94142215173143e-05,
+      "loss": 1.5323,
+      "step": 2226
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4591786861419678,
+      "learning_rate": 4.9413957176843776e-05,
+      "loss": 1.3624,
+      "step": 2227
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4736456871032715,
+      "learning_rate": 4.941369283637325e-05,
+      "loss": 1.4736,
+      "step": 2228
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.741546869277954,
+      "learning_rate": 4.9413428495902724e-05,
+      "loss": 1.4725,
+      "step": 2229
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.764255404472351,
+      "learning_rate": 4.9413164155432204e-05,
+      "loss": 1.5541,
+      "step": 2230
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0397331714630127,
+      "learning_rate": 4.941289981496167e-05,
+      "loss": 1.4632,
+      "step": 2231
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4291046857833862,
+      "learning_rate": 4.941263547449115e-05,
+      "loss": 1.4215,
+      "step": 2232
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4655812978744507,
+      "learning_rate": 4.941237113402062e-05,
+      "loss": 1.5164,
+      "step": 2233
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6057709455490112,
+      "learning_rate": 4.941210679355009e-05,
+      "loss": 1.3987,
+      "step": 2234
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6227375268936157,
+      "learning_rate": 4.941184245307957e-05,
+      "loss": 1.3583,
+      "step": 2235
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8403795957565308,
+      "learning_rate": 4.941157811260904e-05,
+      "loss": 1.4767,
+      "step": 2236
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.227350950241089,
+      "learning_rate": 4.941131377213852e-05,
+      "loss": 1.7063,
+      "step": 2237
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.838769555091858,
+      "learning_rate": 4.9411049431667986e-05,
+      "loss": 1.37,
+      "step": 2238
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5935840606689453,
+      "learning_rate": 4.9410785091197466e-05,
+      "loss": 1.4101,
+      "step": 2239
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.447723388671875,
+      "learning_rate": 4.941052075072694e-05,
+      "loss": 1.3171,
+      "step": 2240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3809573650360107,
+      "learning_rate": 4.941025641025641e-05,
+      "loss": 1.245,
+      "step": 2241
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7944077253341675,
+      "learning_rate": 4.940999206978589e-05,
+      "loss": 1.3818,
+      "step": 2242
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6397761106491089,
+      "learning_rate": 4.940972772931536e-05,
+      "loss": 1.5691,
+      "step": 2243
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5444424152374268,
+      "learning_rate": 4.9409463388844834e-05,
+      "loss": 1.2982,
+      "step": 2244
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.766074776649475,
+      "learning_rate": 4.940919904837431e-05,
+      "loss": 1.4268,
+      "step": 2245
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7646783590316772,
+      "learning_rate": 4.940893470790378e-05,
+      "loss": 1.571,
+      "step": 2246
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3304673433303833,
+      "learning_rate": 4.9408670367433255e-05,
+      "loss": 1.1436,
+      "step": 2247
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.312380313873291,
+      "learning_rate": 4.940840602696273e-05,
+      "loss": 1.2443,
+      "step": 2248
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3655829429626465,
+      "learning_rate": 4.94081416864922e-05,
+      "loss": 1.2238,
+      "step": 2249
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9820133447647095,
+      "learning_rate": 4.940787734602168e-05,
+      "loss": 1.3891,
+      "step": 2250
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.653261423110962,
+      "learning_rate": 4.940761300555115e-05,
+      "loss": 1.4225,
+      "step": 2251
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5320602655410767,
+      "learning_rate": 4.940734866508063e-05,
+      "loss": 1.3469,
+      "step": 2252
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5594828128814697,
+      "learning_rate": 4.9407084324610096e-05,
+      "loss": 1.4535,
+      "step": 2253
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6855921745300293,
+      "learning_rate": 4.9406819984139576e-05,
+      "loss": 1.4432,
+      "step": 2254
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3381190299987793,
+      "learning_rate": 4.940655564366905e-05,
+      "loss": 1.5325,
+      "step": 2255
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.521158218383789,
+      "learning_rate": 4.9406291303198524e-05,
+      "loss": 1.5014,
+      "step": 2256
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.567913055419922,
+      "learning_rate": 4.9406026962728e-05,
+      "loss": 1.4007,
+      "step": 2257
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.649031639099121,
+      "learning_rate": 4.940576262225747e-05,
+      "loss": 1.4985,
+      "step": 2258
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0005736351013184,
+      "learning_rate": 4.9405498281786944e-05,
+      "loss": 3.9076,
+      "step": 2259
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.827807903289795,
+      "learning_rate": 4.940523394131642e-05,
+      "loss": 1.533,
+      "step": 2260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7575260400772095,
+      "learning_rate": 4.940496960084589e-05,
+      "loss": 1.6206,
+      "step": 2261
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2160979509353638,
+      "learning_rate": 4.9404705260375365e-05,
+      "loss": 1.5688,
+      "step": 2262
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5280649662017822,
+      "learning_rate": 4.940444091990484e-05,
+      "loss": 1.2798,
+      "step": 2263
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.417629361152649,
+      "learning_rate": 4.940417657943431e-05,
+      "loss": 1.406,
+      "step": 2264
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5411386489868164,
+      "learning_rate": 4.940391223896379e-05,
+      "loss": 1.4835,
+      "step": 2265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7775455713272095,
+      "learning_rate": 4.940364789849326e-05,
+      "loss": 1.526,
+      "step": 2266
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4209047555923462,
+      "learning_rate": 4.940338355802274e-05,
+      "loss": 1.4785,
+      "step": 2267
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6156156063079834,
+      "learning_rate": 4.9403119217552206e-05,
+      "loss": 1.6491,
+      "step": 2268
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.774096965789795,
+      "learning_rate": 4.940285487708169e-05,
+      "loss": 1.5299,
+      "step": 2269
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8696895837783813,
+      "learning_rate": 4.940259053661116e-05,
+      "loss": 1.461,
+      "step": 2270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7859619855880737,
+      "learning_rate": 4.940232619614063e-05,
+      "loss": 1.5279,
+      "step": 2271
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7771594524383545,
+      "learning_rate": 4.940206185567011e-05,
+      "loss": 1.4298,
+      "step": 2272
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7193577289581299,
+      "learning_rate": 4.9401797515199574e-05,
+      "loss": 2.0005,
+      "step": 2273
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.780691146850586,
+      "learning_rate": 4.9401533174729055e-05,
+      "loss": 1.4272,
+      "step": 2274
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.894639253616333,
+      "learning_rate": 4.940126883425853e-05,
+      "loss": 1.4521,
+      "step": 2275
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7083796262741089,
+      "learning_rate": 4.9401004493788e-05,
+      "loss": 1.4993,
+      "step": 2276
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2143800258636475,
+      "learning_rate": 4.9400740153317475e-05,
+      "loss": 1.6448,
+      "step": 2277
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5046420097351074,
+      "learning_rate": 4.940047581284695e-05,
+      "loss": 1.406,
+      "step": 2278
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.932976722717285,
+      "learning_rate": 4.940021147237642e-05,
+      "loss": 1.4236,
+      "step": 2279
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4379674196243286,
+      "learning_rate": 4.9399947131905896e-05,
+      "loss": 1.5148,
+      "step": 2280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.131671667098999,
+      "learning_rate": 4.939968279143537e-05,
+      "loss": 1.611,
+      "step": 2281
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.509517788887024,
+      "learning_rate": 4.939941845096484e-05,
+      "loss": 1.4047,
+      "step": 2282
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7812248468399048,
+      "learning_rate": 4.939915411049432e-05,
+      "loss": 1.5604,
+      "step": 2283
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4432005882263184,
+      "learning_rate": 4.939888977002379e-05,
+      "loss": 1.3953,
+      "step": 2284
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6380198001861572,
+      "learning_rate": 4.939862542955327e-05,
+      "loss": 1.3126,
+      "step": 2285
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9018352031707764,
+      "learning_rate": 4.939836108908274e-05,
+      "loss": 1.2731,
+      "step": 2286
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5882631540298462,
+      "learning_rate": 4.939809674861222e-05,
+      "loss": 2.2868,
+      "step": 2287
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.525963544845581,
+      "learning_rate": 4.9397832408141685e-05,
+      "loss": 1.3599,
+      "step": 2288
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4691402912139893,
+      "learning_rate": 4.9397568067671165e-05,
+      "loss": 1.3277,
+      "step": 2289
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.781137466430664,
+      "learning_rate": 4.939730372720064e-05,
+      "loss": 1.6099,
+      "step": 2290
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8846286535263062,
+      "learning_rate": 4.939703938673011e-05,
+      "loss": 1.4678,
+      "step": 2291
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.665902018547058,
+      "learning_rate": 4.9396775046259586e-05,
+      "loss": 1.3785,
+      "step": 2292
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6956348419189453,
+      "learning_rate": 4.939651070578906e-05,
+      "loss": 1.6033,
+      "step": 2293
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7793266773223877,
+      "learning_rate": 4.939624636531853e-05,
+      "loss": 1.4821,
+      "step": 2294
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.599941611289978,
+      "learning_rate": 4.9395982024848006e-05,
+      "loss": 1.4703,
+      "step": 2295
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7377068996429443,
+      "learning_rate": 4.939571768437748e-05,
+      "loss": 1.3383,
+      "step": 2296
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5180423259735107,
+      "learning_rate": 4.9395453343906954e-05,
+      "loss": 1.3725,
+      "step": 2297
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.780652642250061,
+      "learning_rate": 4.939518900343643e-05,
+      "loss": 1.4059,
+      "step": 2298
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4120699167251587,
+      "learning_rate": 4.93949246629659e-05,
+      "loss": 1.3599,
+      "step": 2299
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3082680702209473,
+      "learning_rate": 4.939466032249538e-05,
+      "loss": 1.4206,
+      "step": 2300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5319504737854004,
+      "learning_rate": 4.939439598202485e-05,
+      "loss": 1.3159,
+      "step": 2301
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.247488260269165,
+      "learning_rate": 4.939413164155433e-05,
+      "loss": 1.6713,
+      "step": 2302
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.429728627204895,
+      "learning_rate": 4.9393867301083795e-05,
+      "loss": 1.2233,
+      "step": 2303
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9966219663619995,
+      "learning_rate": 4.9393602960613275e-05,
+      "loss": 1.5025,
+      "step": 2304
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4088354110717773,
+      "learning_rate": 4.939333862014275e-05,
+      "loss": 1.3309,
+      "step": 2305
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4995232820510864,
+      "learning_rate": 4.9393074279672216e-05,
+      "loss": 1.4,
+      "step": 2306
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8407751321792603,
+      "learning_rate": 4.9392809939201696e-05,
+      "loss": 1.4836,
+      "step": 2307
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4846382141113281,
+      "learning_rate": 4.939254559873116e-05,
+      "loss": 1.3228,
+      "step": 2308
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4750442504882812,
+      "learning_rate": 4.939228125826064e-05,
+      "loss": 1.4781,
+      "step": 2309
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5376219749450684,
+      "learning_rate": 4.939201691779012e-05,
+      "loss": 1.5541,
+      "step": 2310
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6533149480819702,
+      "learning_rate": 4.939175257731959e-05,
+      "loss": 1.6193,
+      "step": 2311
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8189752101898193,
+      "learning_rate": 4.9391488236849064e-05,
+      "loss": 1.6956,
+      "step": 2312
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.692077398300171,
+      "learning_rate": 4.939122389637854e-05,
+      "loss": 1.4977,
+      "step": 2313
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5560657978057861,
+      "learning_rate": 4.939095955590801e-05,
+      "loss": 1.4973,
+      "step": 2314
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.081965923309326,
+      "learning_rate": 4.939069521543749e-05,
+      "loss": 1.4793,
+      "step": 2315
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5001803636550903,
+      "learning_rate": 4.939043087496696e-05,
+      "loss": 1.2482,
+      "step": 2316
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4019306898117065,
+      "learning_rate": 4.939016653449643e-05,
+      "loss": 1.4347,
+      "step": 2317
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.88246488571167,
+      "learning_rate": 4.9389902194025905e-05,
+      "loss": 1.4003,
+      "step": 2318
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6375524997711182,
+      "learning_rate": 4.938963785355538e-05,
+      "loss": 1.3224,
+      "step": 2319
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.255948543548584,
+      "learning_rate": 4.938937351308486e-05,
+      "loss": 1.2805,
+      "step": 2320
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4010885953903198,
+      "learning_rate": 4.9389109172614326e-05,
+      "loss": 1.4506,
+      "step": 2321
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1174590587615967,
+      "learning_rate": 4.9388844832143806e-05,
+      "loss": 1.4425,
+      "step": 2322
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9042725563049316,
+      "learning_rate": 4.938858049167327e-05,
+      "loss": 1.421,
+      "step": 2323
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.707708477973938,
+      "learning_rate": 4.9388316151202754e-05,
+      "loss": 1.4613,
+      "step": 2324
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.81717848777771,
+      "learning_rate": 4.938805181073223e-05,
+      "loss": 1.6336,
+      "step": 2325
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4562922716140747,
+      "learning_rate": 4.93877874702617e-05,
+      "loss": 1.379,
+      "step": 2326
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.418179988861084,
+      "learning_rate": 4.9387523129791174e-05,
+      "loss": 1.6784,
+      "step": 2327
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7778843641281128,
+      "learning_rate": 4.938725878932065e-05,
+      "loss": 1.3697,
+      "step": 2328
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4380995035171509,
+      "learning_rate": 4.938699444885012e-05,
+      "loss": 1.524,
+      "step": 2329
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.611218452453613,
+      "learning_rate": 4.9386730108379595e-05,
+      "loss": 1.8301,
+      "step": 2330
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.842740774154663,
+      "learning_rate": 4.938646576790907e-05,
+      "loss": 1.4556,
+      "step": 2331
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.866092562675476,
+      "learning_rate": 4.938620142743854e-05,
+      "loss": 1.3859,
+      "step": 2332
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7981314659118652,
+      "learning_rate": 4.9385937086968016e-05,
+      "loss": 1.544,
+      "step": 2333
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6464797258377075,
+      "learning_rate": 4.938567274649749e-05,
+      "loss": 1.4708,
+      "step": 2334
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5154670476913452,
+      "learning_rate": 4.938540840602697e-05,
+      "loss": 1.5394,
+      "step": 2335
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.022327423095703,
+      "learning_rate": 4.9385144065556436e-05,
+      "loss": 1.5535,
+      "step": 2336
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.661717176437378,
+      "learning_rate": 4.938487972508592e-05,
+      "loss": 1.5126,
+      "step": 2337
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0259299278259277,
+      "learning_rate": 4.9384615384615384e-05,
+      "loss": 1.6306,
+      "step": 2338
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.572228193283081,
+      "learning_rate": 4.9384351044144864e-05,
+      "loss": 1.4354,
+      "step": 2339
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7299202680587769,
+      "learning_rate": 4.938408670367434e-05,
+      "loss": 1.483,
+      "step": 2340
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.203014612197876,
+      "learning_rate": 4.938382236320381e-05,
+      "loss": 1.4084,
+      "step": 2341
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6922742128372192,
+      "learning_rate": 4.9383558022733285e-05,
+      "loss": 1.3097,
+      "step": 2342
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.874627947807312,
+      "learning_rate": 4.938329368226275e-05,
+      "loss": 1.3834,
+      "step": 2343
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5135424137115479,
+      "learning_rate": 4.938302934179223e-05,
+      "loss": 1.4862,
+      "step": 2344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.047708034515381,
+      "learning_rate": 4.9382765001321705e-05,
+      "loss": 1.6167,
+      "step": 2345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8067690134048462,
+      "learning_rate": 4.938250066085118e-05,
+      "loss": 1.5432,
+      "step": 2346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8268177509307861,
+      "learning_rate": 4.938223632038065e-05,
+      "loss": 1.5669,
+      "step": 2347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.334329605102539,
+      "learning_rate": 4.9381971979910126e-05,
+      "loss": 2.0007,
+      "step": 2348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.44378662109375,
+      "learning_rate": 4.93817076394396e-05,
+      "loss": 1.6891,
+      "step": 2349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9474676847457886,
+      "learning_rate": 4.938144329896908e-05,
+      "loss": 1.5465,
+      "step": 2350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3577466011047363,
+      "learning_rate": 4.938117895849855e-05,
+      "loss": 1.1966,
+      "step": 2351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.747852087020874,
+      "learning_rate": 4.938091461802802e-05,
+      "loss": 1.5766,
+      "step": 2352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4757788181304932,
+      "learning_rate": 4.9380650277557494e-05,
+      "loss": 1.3046,
+      "step": 2353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5664772987365723,
+      "learning_rate": 4.938038593708697e-05,
+      "loss": 1.3238,
+      "step": 2354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.106505870819092,
+      "learning_rate": 4.938012159661645e-05,
+      "loss": 1.5166,
+      "step": 2355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.749468207359314,
+      "learning_rate": 4.9379857256145915e-05,
+      "loss": 1.3865,
+      "step": 2356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9592474699020386,
+      "learning_rate": 4.9379592915675395e-05,
+      "loss": 1.4182,
+      "step": 2357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5040282011032104,
+      "learning_rate": 4.937932857520486e-05,
+      "loss": 1.4017,
+      "step": 2358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5454840660095215,
+      "learning_rate": 4.937906423473434e-05,
+      "loss": 1.3641,
+      "step": 2359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.382007360458374,
+      "learning_rate": 4.9378799894263816e-05,
+      "loss": 1.4683,
+      "step": 2360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7899610996246338,
+      "learning_rate": 4.937853555379329e-05,
+      "loss": 1.4196,
+      "step": 2361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.127277135848999,
+      "learning_rate": 4.937827121332276e-05,
+      "loss": 1.314,
+      "step": 2362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7370450496673584,
+      "learning_rate": 4.9378006872852236e-05,
+      "loss": 1.3111,
+      "step": 2363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1223154067993164,
+      "learning_rate": 4.937774253238171e-05,
+      "loss": 1.8038,
+      "step": 2364
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4738236665725708,
+      "learning_rate": 4.9377478191911183e-05,
+      "loss": 1.4471,
+      "step": 2365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7869237661361694,
+      "learning_rate": 4.937721385144066e-05,
+      "loss": 1.4036,
+      "step": 2366
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.536539912223816,
+      "learning_rate": 4.937694951097013e-05,
+      "loss": 1.435,
+      "step": 2367
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6385867595672607,
+      "learning_rate": 4.9376685170499604e-05,
+      "loss": 1.4819,
+      "step": 2368
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5523381233215332,
+      "learning_rate": 4.937642083002908e-05,
+      "loss": 1.4245,
+      "step": 2369
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8514865636825562,
+      "learning_rate": 4.937615648955856e-05,
+      "loss": 1.7092,
+      "step": 2370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5022422075271606,
+      "learning_rate": 4.9375892149088025e-05,
+      "loss": 1.4798,
+      "step": 2371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.885557770729065,
+      "learning_rate": 4.9375627808617505e-05,
+      "loss": 1.5498,
+      "step": 2372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7222952842712402,
+      "learning_rate": 4.937536346814697e-05,
+      "loss": 1.4686,
+      "step": 2373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.574792504310608,
+      "learning_rate": 4.937509912767645e-05,
+      "loss": 1.2613,
+      "step": 2374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6948360204696655,
+      "learning_rate": 4.9374834787205926e-05,
+      "loss": 1.4965,
+      "step": 2375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8152841329574585,
+      "learning_rate": 4.93745704467354e-05,
+      "loss": 1.4643,
+      "step": 2376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.494125485420227,
+      "learning_rate": 4.937430610626487e-05,
+      "loss": 1.4934,
+      "step": 2377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5091161727905273,
+      "learning_rate": 4.937404176579434e-05,
+      "loss": 1.3528,
+      "step": 2378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8950613737106323,
+      "learning_rate": 4.937377742532382e-05,
+      "loss": 1.4897,
+      "step": 2379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5004290342330933,
+      "learning_rate": 4.9373513084853294e-05,
+      "loss": 1.3739,
+      "step": 2380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.58848237991333,
+      "learning_rate": 4.937324874438277e-05,
+      "loss": 1.5027,
+      "step": 2381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7720941305160522,
+      "learning_rate": 4.937298440391224e-05,
+      "loss": 1.4026,
+      "step": 2382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8544623851776123,
+      "learning_rate": 4.9372720063441715e-05,
+      "loss": 1.4294,
+      "step": 2383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7314934730529785,
+      "learning_rate": 4.937245572297119e-05,
+      "loss": 1.6276,
+      "step": 2384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8919404745101929,
+      "learning_rate": 4.937219138250067e-05,
+      "loss": 1.5107,
+      "step": 2385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.396770477294922,
+      "learning_rate": 4.9371927042030135e-05,
+      "loss": 1.4019,
+      "step": 2386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9726793766021729,
+      "learning_rate": 4.9371662701559616e-05,
+      "loss": 2.4706,
+      "step": 2387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6846975088119507,
+      "learning_rate": 4.937139836108908e-05,
+      "loss": 1.4213,
+      "step": 2388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4821696281433105,
+      "learning_rate": 4.9371134020618556e-05,
+      "loss": 1.5949,
+      "step": 2389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5442427396774292,
+      "learning_rate": 4.9370869680148036e-05,
+      "loss": 1.5166,
+      "step": 2390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4793864488601685,
+      "learning_rate": 4.93706053396775e-05,
+      "loss": 1.3003,
+      "step": 2391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6566134691238403,
+      "learning_rate": 4.9370340999206983e-05,
+      "loss": 1.4139,
+      "step": 2392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6670430898666382,
+      "learning_rate": 4.937007665873645e-05,
+      "loss": 1.3987,
+      "step": 2393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.539172649383545,
+      "learning_rate": 4.936981231826593e-05,
+      "loss": 1.4896,
+      "step": 2394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.361189603805542,
+      "learning_rate": 4.9369547977795404e-05,
+      "loss": 1.254,
+      "step": 2395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8316339254379272,
+      "learning_rate": 4.936928363732488e-05,
+      "loss": 1.459,
+      "step": 2396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3131470680236816,
+      "learning_rate": 4.936901929685435e-05,
+      "loss": 1.331,
+      "step": 2397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.715828776359558,
+      "learning_rate": 4.9368754956383825e-05,
+      "loss": 1.5573,
+      "step": 2398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.236410617828369,
+      "learning_rate": 4.93684906159133e-05,
+      "loss": 1.329,
+      "step": 2399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7689476013183594,
+      "learning_rate": 4.936822627544277e-05,
+      "loss": 1.4756,
+      "step": 2400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4904770851135254,
+      "learning_rate": 4.9367961934972246e-05,
+      "loss": 1.1857,
+      "step": 2401
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7221384048461914,
+      "learning_rate": 4.936769759450172e-05,
+      "loss": 1.3849,
+      "step": 2402
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.882030725479126,
+      "learning_rate": 4.936743325403119e-05,
+      "loss": 1.3904,
+      "step": 2403
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3419508934020996,
+      "learning_rate": 4.9367168913560666e-05,
+      "loss": 1.6619,
+      "step": 2404
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.800032138824463,
+      "learning_rate": 4.936690457309015e-05,
+      "loss": 1.3256,
+      "step": 2405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8268020153045654,
+      "learning_rate": 4.9366640232619613e-05,
+      "loss": 1.5552,
+      "step": 2406
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.686284303665161,
+      "learning_rate": 4.9366375892149094e-05,
+      "loss": 1.7074,
+      "step": 2407
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.826035499572754,
+      "learning_rate": 4.936611155167856e-05,
+      "loss": 1.436,
+      "step": 2408
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.76041841506958,
+      "learning_rate": 4.936584721120804e-05,
+      "loss": 1.7889,
+      "step": 2409
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5033538341522217,
+      "learning_rate": 4.9365582870737515e-05,
+      "loss": 1.4129,
+      "step": 2410
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8592692613601685,
+      "learning_rate": 4.936531853026699e-05,
+      "loss": 1.5551,
+      "step": 2411
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.037543773651123,
+      "learning_rate": 4.936505418979646e-05,
+      "loss": 1.519,
+      "step": 2412
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5872172117233276,
+      "learning_rate": 4.9364789849325935e-05,
+      "loss": 1.4121,
+      "step": 2413
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.67208731174469,
+      "learning_rate": 4.936452550885541e-05,
+      "loss": 1.4447,
+      "step": 2414
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9387143850326538,
+      "learning_rate": 4.936426116838488e-05,
+      "loss": 1.4624,
+      "step": 2415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.028658390045166,
+      "learning_rate": 4.9363996827914356e-05,
+      "loss": 1.4826,
+      "step": 2416
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7263823747634888,
+      "learning_rate": 4.936373248744383e-05,
+      "loss": 1.4597,
+      "step": 2417
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4357963800430298,
+      "learning_rate": 4.93634681469733e-05,
+      "loss": 1.3792,
+      "step": 2418
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.49993097782135,
+      "learning_rate": 4.936320380650278e-05,
+      "loss": 1.4159,
+      "step": 2419
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5787817239761353,
+      "learning_rate": 4.936293946603226e-05,
+      "loss": 1.3902,
+      "step": 2420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0963032245635986,
+      "learning_rate": 4.9362675125561724e-05,
+      "loss": 1.4766,
+      "step": 2421
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4811574220657349,
+      "learning_rate": 4.9362410785091204e-05,
+      "loss": 1.3404,
+      "step": 2422
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9808846712112427,
+      "learning_rate": 4.936214644462067e-05,
+      "loss": 1.3947,
+      "step": 2423
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4518522024154663,
+      "learning_rate": 4.9361882104150145e-05,
+      "loss": 1.4769,
+      "step": 2424
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.560084581375122,
+      "learning_rate": 4.9361617763679625e-05,
+      "loss": 1.4289,
+      "step": 2425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7291773557662964,
+      "learning_rate": 4.936135342320909e-05,
+      "loss": 1.5422,
+      "step": 2426
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7829111814498901,
+      "learning_rate": 4.936108908273857e-05,
+      "loss": 1.6508,
+      "step": 2427
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7511608600616455,
+      "learning_rate": 4.936082474226804e-05,
+      "loss": 1.5588,
+      "step": 2428
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4403977394104004,
+      "learning_rate": 4.936056040179752e-05,
+      "loss": 1.3258,
+      "step": 2429
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2807862758636475,
+      "learning_rate": 4.936029606132699e-05,
+      "loss": 1.603,
+      "step": 2430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5721107721328735,
+      "learning_rate": 4.9360031720856466e-05,
+      "loss": 1.4034,
+      "step": 2431
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7286064624786377,
+      "learning_rate": 4.935976738038594e-05,
+      "loss": 1.5118,
+      "step": 2432
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2323310375213623,
+      "learning_rate": 4.9359503039915413e-05,
+      "loss": 1.3279,
+      "step": 2433
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.712735891342163,
+      "learning_rate": 4.935923869944489e-05,
+      "loss": 1.3691,
+      "step": 2434
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.695513129234314,
+      "learning_rate": 4.935897435897436e-05,
+      "loss": 1.3424,
+      "step": 2435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.77443265914917,
+      "learning_rate": 4.9358710018503834e-05,
+      "loss": 1.9314,
+      "step": 2436
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.648539662361145,
+      "learning_rate": 4.935844567803331e-05,
+      "loss": 1.4429,
+      "step": 2437
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5863761901855469,
+      "learning_rate": 4.935818133756278e-05,
+      "loss": 1.4789,
+      "step": 2438
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8181358575820923,
+      "learning_rate": 4.9357916997092255e-05,
+      "loss": 1.5613,
+      "step": 2439
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0188419818878174,
+      "learning_rate": 4.9357652656621735e-05,
+      "loss": 1.3271,
+      "step": 2440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3554275035858154,
+      "learning_rate": 4.93573883161512e-05,
+      "loss": 1.2874,
+      "step": 2441
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.555167555809021,
+      "learning_rate": 4.935712397568068e-05,
+      "loss": 2.5132,
+      "step": 2442
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9438148736953735,
+      "learning_rate": 4.935685963521015e-05,
+      "loss": 1.406,
+      "step": 2443
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.946184754371643,
+      "learning_rate": 4.935659529473963e-05,
+      "loss": 1.3386,
+      "step": 2444
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5025683641433716,
+      "learning_rate": 4.93563309542691e-05,
+      "loss": 2.5231,
+      "step": 2445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.248850107192993,
+      "learning_rate": 4.935606661379858e-05,
+      "loss": 1.5518,
+      "step": 2446
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5831654071807861,
+      "learning_rate": 4.935580227332805e-05,
+      "loss": 1.2772,
+      "step": 2447
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5145089626312256,
+      "learning_rate": 4.9355537932857524e-05,
+      "loss": 1.556,
+      "step": 2448
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7781612873077393,
+      "learning_rate": 4.9355273592387e-05,
+      "loss": 1.5176,
+      "step": 2449
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6996303796768188,
+      "learning_rate": 4.935500925191647e-05,
+      "loss": 1.4003,
+      "step": 2450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3697829246520996,
+      "learning_rate": 4.9354744911445944e-05,
+      "loss": 1.7504,
+      "step": 2451
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.392005443572998,
+      "learning_rate": 4.935448057097542e-05,
+      "loss": 1.3357,
+      "step": 2452
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.092898368835449,
+      "learning_rate": 4.935421623050489e-05,
+      "loss": 1.5416,
+      "step": 2453
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6516717672348022,
+      "learning_rate": 4.9353951890034365e-05,
+      "loss": 1.3676,
+      "step": 2454
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5954350233078003,
+      "learning_rate": 4.9353687549563846e-05,
+      "loss": 1.558,
+      "step": 2455
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4154949188232422,
+      "learning_rate": 4.935342320909331e-05,
+      "loss": 1.4256,
+      "step": 2456
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3137110471725464,
+      "learning_rate": 4.935315886862279e-05,
+      "loss": 1.356,
+      "step": 2457
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5134217739105225,
+      "learning_rate": 4.935289452815226e-05,
+      "loss": 1.335,
+      "step": 2458
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2668488025665283,
+      "learning_rate": 4.935263018768174e-05,
+      "loss": 1.6238,
+      "step": 2459
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4393428564071655,
+      "learning_rate": 4.935236584721121e-05,
+      "loss": 1.2546,
+      "step": 2460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4637144804000854,
+      "learning_rate": 4.935210150674068e-05,
+      "loss": 1.4516,
+      "step": 2461
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4443246126174927,
+      "learning_rate": 4.935183716627016e-05,
+      "loss": 1.3623,
+      "step": 2462
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5822737216949463,
+      "learning_rate": 4.935157282579963e-05,
+      "loss": 1.4509,
+      "step": 2463
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.834412932395935,
+      "learning_rate": 4.935130848532911e-05,
+      "loss": 1.4784,
+      "step": 2464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9873172044754028,
+      "learning_rate": 4.935104414485858e-05,
+      "loss": 1.5835,
+      "step": 2465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7813200950622559,
+      "learning_rate": 4.9350779804388055e-05,
+      "loss": 1.5462,
+      "step": 2466
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.94339919090271,
+      "learning_rate": 4.935051546391753e-05,
+      "loss": 1.3722,
+      "step": 2467
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.852325677871704,
+      "learning_rate": 4.9350251123447e-05,
+      "loss": 1.5978,
+      "step": 2468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5803498029708862,
+      "learning_rate": 4.9349986782976476e-05,
+      "loss": 1.434,
+      "step": 2469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2676278352737427,
+      "learning_rate": 4.934972244250595e-05,
+      "loss": 1.3663,
+      "step": 2470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4079687595367432,
+      "learning_rate": 4.934945810203542e-05,
+      "loss": 1.3785,
+      "step": 2471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8618310689926147,
+      "learning_rate": 4.9349193761564896e-05,
+      "loss": 1.3932,
+      "step": 2472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7892003059387207,
+      "learning_rate": 4.934892942109437e-05,
+      "loss": 1.4496,
+      "step": 2473
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.357029438018799,
+      "learning_rate": 4.934866508062384e-05,
+      "loss": 1.6165,
+      "step": 2474
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5947681665420532,
+      "learning_rate": 4.9348400740153324e-05,
+      "loss": 1.4495,
+      "step": 2475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9325721263885498,
+      "learning_rate": 4.934813639968279e-05,
+      "loss": 1.5673,
+      "step": 2476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8947330713272095,
+      "learning_rate": 4.934787205921227e-05,
+      "loss": 1.4747,
+      "step": 2477
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.695472240447998,
+      "learning_rate": 4.934760771874174e-05,
+      "loss": 1.4743,
+      "step": 2478
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6457475423812866,
+      "learning_rate": 4.934734337827122e-05,
+      "loss": 1.4162,
+      "step": 2479
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3956620693206787,
+      "learning_rate": 4.934707903780069e-05,
+      "loss": 2.1544,
+      "step": 2480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8883674144744873,
+      "learning_rate": 4.9346814697330165e-05,
+      "loss": 1.543,
+      "step": 2481
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7279736995697021,
+      "learning_rate": 4.934655035685964e-05,
+      "loss": 1.4069,
+      "step": 2482
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6493021249771118,
+      "learning_rate": 4.934628601638911e-05,
+      "loss": 1.5576,
+      "step": 2483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7916574478149414,
+      "learning_rate": 4.9346021675918586e-05,
+      "loss": 1.5453,
+      "step": 2484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5575659275054932,
+      "learning_rate": 4.934575733544806e-05,
+      "loss": 1.5546,
+      "step": 2485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6311609745025635,
+      "learning_rate": 4.934549299497753e-05,
+      "loss": 1.4947,
+      "step": 2486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9200974702835083,
+      "learning_rate": 4.9345228654507007e-05,
+      "loss": 1.5567,
+      "step": 2487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7935091257095337,
+      "learning_rate": 4.934496431403648e-05,
+      "loss": 1.5087,
+      "step": 2488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2324535846710205,
+      "learning_rate": 4.9344699973565954e-05,
+      "loss": 1.6002,
+      "step": 2489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4483072757720947,
+      "learning_rate": 4.9344435633095434e-05,
+      "loss": 1.3008,
+      "step": 2490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.843936800956726,
+      "learning_rate": 4.93441712926249e-05,
+      "loss": 1.4899,
+      "step": 2491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9429736137390137,
+      "learning_rate": 4.934390695215438e-05,
+      "loss": 1.3964,
+      "step": 2492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.492895245552063,
+      "learning_rate": 4.934364261168385e-05,
+      "loss": 1.5067,
+      "step": 2493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7376607656478882,
+      "learning_rate": 4.934337827121333e-05,
+      "loss": 1.3531,
+      "step": 2494
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.047926187515259,
+      "learning_rate": 4.93431139307428e-05,
+      "loss": 1.3738,
+      "step": 2495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6065254211425781,
+      "learning_rate": 4.934284959027227e-05,
+      "loss": 1.3717,
+      "step": 2496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6963759660720825,
+      "learning_rate": 4.934258524980175e-05,
+      "loss": 1.5242,
+      "step": 2497
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8830819129943848,
+      "learning_rate": 4.9342320909331216e-05,
+      "loss": 1.2895,
+      "step": 2498
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7731037139892578,
+      "learning_rate": 4.9342056568860696e-05,
+      "loss": 1.529,
+      "step": 2499
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7736376523971558,
+      "learning_rate": 4.934179222839017e-05,
+      "loss": 1.608,
+      "step": 2500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6290074586868286,
+      "learning_rate": 4.934152788791964e-05,
+      "loss": 1.3643,
+      "step": 2501
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4773492813110352,
+      "learning_rate": 4.934126354744912e-05,
+      "loss": 1.2985,
+      "step": 2502
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3611364364624023,
+      "learning_rate": 4.934099920697859e-05,
+      "loss": 1.3573,
+      "step": 2503
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7505794763565063,
+      "learning_rate": 4.9340734866508064e-05,
+      "loss": 1.4997,
+      "step": 2504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7649352550506592,
+      "learning_rate": 4.9340470526037544e-05,
+      "loss": 1.4299,
+      "step": 2505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.811025857925415,
+      "learning_rate": 4.934020618556701e-05,
+      "loss": 1.4919,
+      "step": 2506
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6091090440750122,
+      "learning_rate": 4.9339941845096485e-05,
+      "loss": 1.3229,
+      "step": 2507
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.736756443977356,
+      "learning_rate": 4.933967750462596e-05,
+      "loss": 1.5237,
+      "step": 2508
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.389913558959961,
+      "learning_rate": 4.933941316415543e-05,
+      "loss": 1.4543,
+      "step": 2509
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6648410558700562,
+      "learning_rate": 4.933914882368491e-05,
+      "loss": 1.4868,
+      "step": 2510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1838812828063965,
+      "learning_rate": 4.933888448321438e-05,
+      "loss": 1.5691,
+      "step": 2511
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5637959241867065,
+      "learning_rate": 4.933862014274386e-05,
+      "loss": 1.4126,
+      "step": 2512
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.952088475227356,
+      "learning_rate": 4.9338355802273326e-05,
+      "loss": 1.3893,
+      "step": 2513
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6216070652008057,
+      "learning_rate": 4.9338091461802807e-05,
+      "loss": 1.396,
+      "step": 2514
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.481932282447815,
+      "learning_rate": 4.933782712133228e-05,
+      "loss": 1.3529,
+      "step": 2515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.389082193374634,
+      "learning_rate": 4.9337562780861754e-05,
+      "loss": 1.568,
+      "step": 2516
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9801472425460815,
+      "learning_rate": 4.933729844039123e-05,
+      "loss": 1.3813,
+      "step": 2517
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4490299224853516,
+      "learning_rate": 4.93370340999207e-05,
+      "loss": 1.5272,
+      "step": 2518
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4942859411239624,
+      "learning_rate": 4.9336769759450174e-05,
+      "loss": 1.3939,
+      "step": 2519
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8432228565216064,
+      "learning_rate": 4.933650541897965e-05,
+      "loss": 1.5363,
+      "step": 2520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5745909214019775,
+      "learning_rate": 4.933624107850912e-05,
+      "loss": 1.4685,
+      "step": 2521
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6000694036483765,
+      "learning_rate": 4.9335976738038595e-05,
+      "loss": 2.3137,
+      "step": 2522
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6651588678359985,
+      "learning_rate": 4.933571239756807e-05,
+      "loss": 1.4241,
+      "step": 2523
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7461001873016357,
+      "learning_rate": 4.933544805709754e-05,
+      "loss": 1.4987,
+      "step": 2524
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6552432775497437,
+      "learning_rate": 4.933518371662702e-05,
+      "loss": 1.4343,
+      "step": 2525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.854166030883789,
+      "learning_rate": 4.933491937615649e-05,
+      "loss": 1.4251,
+      "step": 2526
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5349853038787842,
+      "learning_rate": 4.933465503568597e-05,
+      "loss": 1.2886,
+      "step": 2527
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4321064949035645,
+      "learning_rate": 4.9334390695215437e-05,
+      "loss": 1.3352,
+      "step": 2528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.750193476676941,
+      "learning_rate": 4.933412635474492e-05,
+      "loss": 1.387,
+      "step": 2529
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.632355809211731,
+      "learning_rate": 4.933386201427439e-05,
+      "loss": 1.3397,
+      "step": 2530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5971956253051758,
+      "learning_rate": 4.9333597673803864e-05,
+      "loss": 1.3191,
+      "step": 2531
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6321008205413818,
+      "learning_rate": 4.933333333333334e-05,
+      "loss": 1.355,
+      "step": 2532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0629541873931885,
+      "learning_rate": 4.9333068992862804e-05,
+      "loss": 1.3837,
+      "step": 2533
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9321011304855347,
+      "learning_rate": 4.9332804652392285e-05,
+      "loss": 1.2453,
+      "step": 2534
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5449974536895752,
+      "learning_rate": 4.933254031192176e-05,
+      "loss": 1.3288,
+      "step": 2535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.231966257095337,
+      "learning_rate": 4.933227597145123e-05,
+      "loss": 1.5703,
+      "step": 2536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5557661056518555,
+      "learning_rate": 4.9332011630980705e-05,
+      "loss": 1.6124,
+      "step": 2537
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2233707904815674,
+      "learning_rate": 4.933174729051018e-05,
+      "loss": 1.5487,
+      "step": 2538
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5379153490066528,
+      "learning_rate": 4.933148295003965e-05,
+      "loss": 1.2759,
+      "step": 2539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6547679901123047,
+      "learning_rate": 4.933121860956913e-05,
+      "loss": 1.3844,
+      "step": 2540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4224644899368286,
+      "learning_rate": 4.93309542690986e-05,
+      "loss": 1.3401,
+      "step": 2541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7587931156158447,
+      "learning_rate": 4.933068992862807e-05,
+      "loss": 1.376,
+      "step": 2542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4771437644958496,
+      "learning_rate": 4.933042558815755e-05,
+      "loss": 1.4334,
+      "step": 2543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1550536155700684,
+      "learning_rate": 4.933016124768702e-05,
+      "loss": 1.6507,
+      "step": 2544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.535224437713623,
+      "learning_rate": 4.93298969072165e-05,
+      "loss": 1.4195,
+      "step": 2545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.201075553894043,
+      "learning_rate": 4.932963256674597e-05,
+      "loss": 1.5509,
+      "step": 2546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7259066104888916,
+      "learning_rate": 4.932936822627545e-05,
+      "loss": 1.5668,
+      "step": 2547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8379123210906982,
+      "learning_rate": 4.9329103885804915e-05,
+      "loss": 1.2651,
+      "step": 2548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4198787212371826,
+      "learning_rate": 4.9328839545334395e-05,
+      "loss": 1.4567,
+      "step": 2549
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6350096464157104,
+      "learning_rate": 4.932857520486387e-05,
+      "loss": 1.4193,
+      "step": 2550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5922600030899048,
+      "learning_rate": 4.932831086439334e-05,
+      "loss": 1.4443,
+      "step": 2551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5865031480789185,
+      "learning_rate": 4.9328046523922816e-05,
+      "loss": 1.3304,
+      "step": 2552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5146315097808838,
+      "learning_rate": 4.932778218345229e-05,
+      "loss": 1.3904,
+      "step": 2553
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7563912868499756,
+      "learning_rate": 4.932751784298176e-05,
+      "loss": 3.2323,
+      "step": 2554
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3881359100341797,
+      "learning_rate": 4.9327253502511237e-05,
+      "loss": 1.5072,
+      "step": 2555
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.51956307888031,
+      "learning_rate": 4.932698916204071e-05,
+      "loss": 1.2903,
+      "step": 2556
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4577417373657227,
+      "learning_rate": 4.9326724821570184e-05,
+      "loss": 1.7638,
+      "step": 2557
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7300041913986206,
+      "learning_rate": 4.932646048109966e-05,
+      "loss": 1.4949,
+      "step": 2558
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4095176458358765,
+      "learning_rate": 4.932619614062913e-05,
+      "loss": 1.362,
+      "step": 2559
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7945948839187622,
+      "learning_rate": 4.932593180015861e-05,
+      "loss": 1.3117,
+      "step": 2560
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.890825629234314,
+      "learning_rate": 4.932566745968808e-05,
+      "loss": 1.3323,
+      "step": 2561
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9443998336791992,
+      "learning_rate": 4.932540311921756e-05,
+      "loss": 1.4265,
+      "step": 2562
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0552046298980713,
+      "learning_rate": 4.9325138778747025e-05,
+      "loss": 1.4211,
+      "step": 2563
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.713041067123413,
+      "learning_rate": 4.9324874438276505e-05,
+      "loss": 1.4382,
+      "step": 2564
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5778248310089111,
+      "learning_rate": 4.932461009780598e-05,
+      "loss": 1.5278,
+      "step": 2565
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9728398323059082,
+      "learning_rate": 4.932434575733545e-05,
+      "loss": 1.5082,
+      "step": 2566
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.496476411819458,
+      "learning_rate": 4.9324081416864926e-05,
+      "loss": 1.3893,
+      "step": 2567
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.676997423171997,
+      "learning_rate": 4.932381707639439e-05,
+      "loss": 1.4032,
+      "step": 2568
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4913362264633179,
+      "learning_rate": 4.932355273592387e-05,
+      "loss": 1.5092,
+      "step": 2569
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8404059410095215,
+      "learning_rate": 4.932328839545335e-05,
+      "loss": 1.3939,
+      "step": 2570
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.274901270866394,
+      "learning_rate": 4.932302405498282e-05,
+      "loss": 1.5177,
+      "step": 2571
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.889642596244812,
+      "learning_rate": 4.9322759714512294e-05,
+      "loss": 1.4386,
+      "step": 2572
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.693866491317749,
+      "learning_rate": 4.932249537404177e-05,
+      "loss": 2.4596,
+      "step": 2573
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.718110203742981,
+      "learning_rate": 4.932223103357124e-05,
+      "loss": 1.5352,
+      "step": 2574
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8123146295547485,
+      "learning_rate": 4.932196669310072e-05,
+      "loss": 1.346,
+      "step": 2575
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9278711080551147,
+      "learning_rate": 4.932170235263019e-05,
+      "loss": 1.4104,
+      "step": 2576
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.035513162612915,
+      "learning_rate": 4.932143801215967e-05,
+      "loss": 1.5067,
+      "step": 2577
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.379963755607605,
+      "learning_rate": 4.9321173671689135e-05,
+      "loss": 1.3781,
+      "step": 2578
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8476982116699219,
+      "learning_rate": 4.932090933121861e-05,
+      "loss": 1.4104,
+      "step": 2579
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3164572715759277,
+      "learning_rate": 4.932064499074809e-05,
+      "loss": 1.2452,
+      "step": 2580
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7981538772583008,
+      "learning_rate": 4.9320380650277556e-05,
+      "loss": 1.5135,
+      "step": 2581
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3418692350387573,
+      "learning_rate": 4.9320116309807036e-05,
+      "loss": 1.3962,
+      "step": 2582
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6116950511932373,
+      "learning_rate": 4.93198519693365e-05,
+      "loss": 1.321,
+      "step": 2583
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4402329921722412,
+      "learning_rate": 4.9319587628865984e-05,
+      "loss": 1.4527,
+      "step": 2584
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6841351985931396,
+      "learning_rate": 4.931932328839546e-05,
+      "loss": 1.5819,
+      "step": 2585
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7440420389175415,
+      "learning_rate": 4.931905894792493e-05,
+      "loss": 1.5349,
+      "step": 2586
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6065852642059326,
+      "learning_rate": 4.9318794607454404e-05,
+      "loss": 1.4461,
+      "step": 2587
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.867735743522644,
+      "learning_rate": 4.931853026698388e-05,
+      "loss": 1.5309,
+      "step": 2588
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4005944728851318,
+      "learning_rate": 4.931826592651335e-05,
+      "loss": 1.341,
+      "step": 2589
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.644071340560913,
+      "learning_rate": 4.9318001586042825e-05,
+      "loss": 1.4234,
+      "step": 2590
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.443773865699768,
+      "learning_rate": 4.93177372455723e-05,
+      "loss": 1.2781,
+      "step": 2591
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0869381427764893,
+      "learning_rate": 4.931747290510177e-05,
+      "loss": 1.606,
+      "step": 2592
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6904364824295044,
+      "learning_rate": 4.9317208564631246e-05,
+      "loss": 1.5036,
+      "step": 2593
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.896661639213562,
+      "learning_rate": 4.931694422416072e-05,
+      "loss": 1.3709,
+      "step": 2594
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0699963569641113,
+      "learning_rate": 4.93166798836902e-05,
+      "loss": 1.4523,
+      "step": 2595
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8072854280471802,
+      "learning_rate": 4.9316415543219666e-05,
+      "loss": 1.5935,
+      "step": 2596
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.534233808517456,
+      "learning_rate": 4.931615120274915e-05,
+      "loss": 1.2914,
+      "step": 2597
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.222546339035034,
+      "learning_rate": 4.9315886862278614e-05,
+      "loss": 1.6006,
+      "step": 2598
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7245173454284668,
+      "learning_rate": 4.9315622521808094e-05,
+      "loss": 1.2704,
+      "step": 2599
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7385029792785645,
+      "learning_rate": 4.931535818133757e-05,
+      "loss": 1.5667,
+      "step": 2600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6196516752243042,
+      "learning_rate": 4.931509384086704e-05,
+      "loss": 1.6003,
+      "step": 2601
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1900652647018433,
+      "learning_rate": 4.9314829500396515e-05,
+      "loss": 1.8619,
+      "step": 2602
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4521243572235107,
+      "learning_rate": 4.931456515992599e-05,
+      "loss": 1.3434,
+      "step": 2603
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6729952096939087,
+      "learning_rate": 4.931430081945546e-05,
+      "loss": 1.4596,
+      "step": 2604
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.708075761795044,
+      "learning_rate": 4.9314036478984935e-05,
+      "loss": 1.4146,
+      "step": 2605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9720501899719238,
+      "learning_rate": 4.931377213851441e-05,
+      "loss": 1.3415,
+      "step": 2606
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4946736097335815,
+      "learning_rate": 4.931350779804388e-05,
+      "loss": 1.4756,
+      "step": 2607
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.050600290298462,
+      "learning_rate": 4.9313243457573356e-05,
+      "loss": 1.6833,
+      "step": 2608
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4357421398162842,
+      "learning_rate": 4.931297911710283e-05,
+      "loss": 1.416,
+      "step": 2609
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.398271083831787,
+      "learning_rate": 4.931271477663231e-05,
+      "loss": 1.259,
+      "step": 2610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.654370665550232,
+      "learning_rate": 4.931245043616178e-05,
+      "loss": 1.5739,
+      "step": 2611
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3777731657028198,
+      "learning_rate": 4.931218609569126e-05,
+      "loss": 1.3652,
+      "step": 2612
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7449884414672852,
+      "learning_rate": 4.9311921755220724e-05,
+      "loss": 2.7371,
+      "step": 2613
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.421327829360962,
+      "learning_rate": 4.93116574147502e-05,
+      "loss": 1.3256,
+      "step": 2614
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6311126947402954,
+      "learning_rate": 4.931139307427968e-05,
+      "loss": 1.2523,
+      "step": 2615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4299547672271729,
+      "learning_rate": 4.9311128733809145e-05,
+      "loss": 1.3461,
+      "step": 2616
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.56879723072052,
+      "learning_rate": 4.9310864393338625e-05,
+      "loss": 1.4114,
+      "step": 2617
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6654942035675049,
+      "learning_rate": 4.931060005286809e-05,
+      "loss": 1.5534,
+      "step": 2618
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9292757511138916,
+      "learning_rate": 4.931033571239757e-05,
+      "loss": 1.4767,
+      "step": 2619
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0756030082702637,
+      "learning_rate": 4.9310071371927046e-05,
+      "loss": 1.5694,
+      "step": 2620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.87142014503479,
+      "learning_rate": 4.930980703145652e-05,
+      "loss": 1.4365,
+      "step": 2621
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4308074712753296,
+      "learning_rate": 4.930954269098599e-05,
+      "loss": 1.3382,
+      "step": 2622
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3024938106536865,
+      "learning_rate": 4.9309278350515466e-05,
+      "loss": 1.3467,
+      "step": 2623
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.58993661403656,
+      "learning_rate": 4.930901401004494e-05,
+      "loss": 1.2993,
+      "step": 2624
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.274555206298828,
+      "learning_rate": 4.9308749669574414e-05,
+      "loss": 1.4688,
+      "step": 2625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3290059566497803,
+      "learning_rate": 4.930848532910389e-05,
+      "loss": 1.6106,
+      "step": 2626
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5617573261260986,
+      "learning_rate": 4.930822098863336e-05,
+      "loss": 1.3978,
+      "step": 2627
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7416939735412598,
+      "learning_rate": 4.9307956648162834e-05,
+      "loss": 1.5047,
+      "step": 2628
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5923374891281128,
+      "learning_rate": 4.930769230769231e-05,
+      "loss": 1.4963,
+      "step": 2629
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4337239265441895,
+      "learning_rate": 4.930742796722179e-05,
+      "loss": 1.835,
+      "step": 2630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.596455454826355,
+      "learning_rate": 4.9307163626751255e-05,
+      "loss": 1.4102,
+      "step": 2631
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.740401268005371,
+      "learning_rate": 4.9306899286280735e-05,
+      "loss": 1.5594,
+      "step": 2632
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8403197526931763,
+      "learning_rate": 4.93066349458102e-05,
+      "loss": 1.5509,
+      "step": 2633
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9447993040084839,
+      "learning_rate": 4.930637060533968e-05,
+      "loss": 1.5905,
+      "step": 2634
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5165345668792725,
+      "learning_rate": 4.9306106264869156e-05,
+      "loss": 1.2023,
+      "step": 2635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6497242450714111,
+      "learning_rate": 4.930584192439863e-05,
+      "loss": 1.5789,
+      "step": 2636
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7731621265411377,
+      "learning_rate": 4.93055775839281e-05,
+      "loss": 1.4642,
+      "step": 2637
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2415149211883545,
+      "learning_rate": 4.930531324345758e-05,
+      "loss": 1.3091,
+      "step": 2638
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5333467721939087,
+      "learning_rate": 4.930504890298705e-05,
+      "loss": 1.3586,
+      "step": 2639
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5875195264816284,
+      "learning_rate": 4.9304784562516524e-05,
+      "loss": 1.4764,
+      "step": 2640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.872005581855774,
+      "learning_rate": 4.9304520222046e-05,
+      "loss": 1.64,
+      "step": 2641
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4437196254730225,
+      "learning_rate": 4.930425588157547e-05,
+      "loss": 1.3744,
+      "step": 2642
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4378817081451416,
+      "learning_rate": 4.9303991541104945e-05,
+      "loss": 1.3808,
+      "step": 2643
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.592041254043579,
+      "learning_rate": 4.930372720063442e-05,
+      "loss": 1.7003,
+      "step": 2644
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5149617195129395,
+      "learning_rate": 4.93034628601639e-05,
+      "loss": 1.3525,
+      "step": 2645
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9863957166671753,
+      "learning_rate": 4.9303198519693365e-05,
+      "loss": 1.53,
+      "step": 2646
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6376615762710571,
+      "learning_rate": 4.9302934179222846e-05,
+      "loss": 1.347,
+      "step": 2647
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4125152826309204,
+      "learning_rate": 4.930266983875231e-05,
+      "loss": 1.2416,
+      "step": 2648
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.938299298286438,
+      "learning_rate": 4.930240549828179e-05,
+      "loss": 1.5191,
+      "step": 2649
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6634442806243896,
+      "learning_rate": 4.9302141157811266e-05,
+      "loss": 1.4438,
+      "step": 2650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8568251132965088,
+      "learning_rate": 4.930187681734073e-05,
+      "loss": 1.5521,
+      "step": 2651
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7931673526763916,
+      "learning_rate": 4.9301612476870214e-05,
+      "loss": 1.2971,
+      "step": 2652
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7395511865615845,
+      "learning_rate": 4.930134813639968e-05,
+      "loss": 1.5296,
+      "step": 2653
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9326555728912354,
+      "learning_rate": 4.930108379592916e-05,
+      "loss": 1.4759,
+      "step": 2654
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.559638500213623,
+      "learning_rate": 4.9300819455458634e-05,
+      "loss": 1.537,
+      "step": 2655
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.219599485397339,
+      "learning_rate": 4.930055511498811e-05,
+      "loss": 1.4938,
+      "step": 2656
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5035227537155151,
+      "learning_rate": 4.930029077451758e-05,
+      "loss": 1.2978,
+      "step": 2657
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.60288405418396,
+      "learning_rate": 4.9300026434047055e-05,
+      "loss": 1.7114,
+      "step": 2658
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9627552032470703,
+      "learning_rate": 4.929976209357653e-05,
+      "loss": 1.3719,
+      "step": 2659
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6352475881576538,
+      "learning_rate": 4.9299497753106e-05,
+      "loss": 1.2557,
+      "step": 2660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0338969230651855,
+      "learning_rate": 4.9299233412635476e-05,
+      "loss": 1.4312,
+      "step": 2661
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.747972011566162,
+      "learning_rate": 4.929896907216495e-05,
+      "loss": 1.4632,
+      "step": 2662
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0859179496765137,
+      "learning_rate": 4.929870473169442e-05,
+      "loss": 1.4356,
+      "step": 2663
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4568469524383545,
+      "learning_rate": 4.9298440391223896e-05,
+      "loss": 1.3237,
+      "step": 2664
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8511412143707275,
+      "learning_rate": 4.929817605075338e-05,
+      "loss": 1.3865,
+      "step": 2665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.44308602809906,
+      "learning_rate": 4.9297911710282844e-05,
+      "loss": 1.3627,
+      "step": 2666
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7719011306762695,
+      "learning_rate": 4.9297647369812324e-05,
+      "loss": 2.0318,
+      "step": 2667
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1529409885406494,
+      "learning_rate": 4.929738302934179e-05,
+      "loss": 1.4575,
+      "step": 2668
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6573741436004639,
+      "learning_rate": 4.929711868887127e-05,
+      "loss": 1.4682,
+      "step": 2669
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7132002115249634,
+      "learning_rate": 4.9296854348400745e-05,
+      "loss": 1.3421,
+      "step": 2670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5792018175125122,
+      "learning_rate": 4.929659000793022e-05,
+      "loss": 1.8775,
+      "step": 2671
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6465085744857788,
+      "learning_rate": 4.929632566745969e-05,
+      "loss": 1.3341,
+      "step": 2672
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4723743200302124,
+      "learning_rate": 4.9296061326989165e-05,
+      "loss": 1.443,
+      "step": 2673
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.909319281578064,
+      "learning_rate": 4.929579698651864e-05,
+      "loss": 1.3724,
+      "step": 2674
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7772451639175415,
+      "learning_rate": 4.929553264604811e-05,
+      "loss": 1.3921,
+      "step": 2675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7596421241760254,
+      "learning_rate": 4.9295268305577586e-05,
+      "loss": 1.4144,
+      "step": 2676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4341310262680054,
+      "learning_rate": 4.929500396510706e-05,
+      "loss": 2.4459,
+      "step": 2677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.804653525352478,
+      "learning_rate": 4.929473962463653e-05,
+      "loss": 1.3996,
+      "step": 2678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3521728515625,
+      "learning_rate": 4.929447528416601e-05,
+      "loss": 1.2909,
+      "step": 2679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6442246437072754,
+      "learning_rate": 4.929421094369549e-05,
+      "loss": 1.439,
+      "step": 2680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5506179332733154,
+      "learning_rate": 4.9293946603224954e-05,
+      "loss": 1.3617,
+      "step": 2681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4139620065689087,
+      "learning_rate": 4.9293682262754434e-05,
+      "loss": 2.3219,
+      "step": 2682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.46586537361145,
+      "learning_rate": 4.92934179222839e-05,
+      "loss": 1.4819,
+      "step": 2683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.680657982826233,
+      "learning_rate": 4.929315358181338e-05,
+      "loss": 1.366,
+      "step": 2684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.419192314147949,
+      "learning_rate": 4.9292889241342855e-05,
+      "loss": 1.3439,
+      "step": 2685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0687263011932373,
+      "learning_rate": 4.929262490087232e-05,
+      "loss": 2.6284,
+      "step": 2686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6514066457748413,
+      "learning_rate": 4.92923605604018e-05,
+      "loss": 1.5053,
+      "step": 2687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8187365531921387,
+      "learning_rate": 4.929209621993127e-05,
+      "loss": 1.4116,
+      "step": 2688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9590576887130737,
+      "learning_rate": 4.929183187946075e-05,
+      "loss": 1.5203,
+      "step": 2689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.030620574951172,
+      "learning_rate": 4.929156753899022e-05,
+      "loss": 1.2171,
+      "step": 2690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5302257537841797,
+      "learning_rate": 4.9291303198519696e-05,
+      "loss": 1.409,
+      "step": 2691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.014828681945801,
+      "learning_rate": 4.929103885804917e-05,
+      "loss": 1.5347,
+      "step": 2692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7564088106155396,
+      "learning_rate": 4.9290774517578643e-05,
+      "loss": 1.5234,
+      "step": 2693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7427843809127808,
+      "learning_rate": 4.929051017710812e-05,
+      "loss": 1.2636,
+      "step": 2694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9185667037963867,
+      "learning_rate": 4.92902458366376e-05,
+      "loss": 1.7086,
+      "step": 2695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2490100860595703,
+      "learning_rate": 4.9289981496167064e-05,
+      "loss": 1.4622,
+      "step": 2696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8235183954238892,
+      "learning_rate": 4.928971715569654e-05,
+      "loss": 1.5331,
+      "step": 2697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6290972232818604,
+      "learning_rate": 4.928945281522601e-05,
+      "loss": 1.3818,
+      "step": 2698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5711686611175537,
+      "learning_rate": 4.9289188474755485e-05,
+      "loss": 1.4374,
+      "step": 2699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.964842438697815,
+      "learning_rate": 4.9288924134284965e-05,
+      "loss": 1.3278,
+      "step": 2700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.770525574684143,
+      "learning_rate": 4.928865979381443e-05,
+      "loss": 1.5938,
+      "step": 2701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6944955587387085,
+      "learning_rate": 4.928839545334391e-05,
+      "loss": 1.456,
+      "step": 2702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4455058574676514,
+      "learning_rate": 4.928813111287338e-05,
+      "loss": 1.3663,
+      "step": 2703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6356016397476196,
+      "learning_rate": 4.928786677240286e-05,
+      "loss": 1.4699,
+      "step": 2704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6325865983963013,
+      "learning_rate": 4.928760243193233e-05,
+      "loss": 1.5538,
+      "step": 2705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4993972778320312,
+      "learning_rate": 4.928733809146181e-05,
+      "loss": 1.2585,
+      "step": 2706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7665947675704956,
+      "learning_rate": 4.928707375099128e-05,
+      "loss": 1.3478,
+      "step": 2707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6483653783798218,
+      "learning_rate": 4.9286809410520754e-05,
+      "loss": 1.3102,
+      "step": 2708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8875410556793213,
+      "learning_rate": 4.928654507005023e-05,
+      "loss": 1.5661,
+      "step": 2709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6641933917999268,
+      "learning_rate": 4.92862807295797e-05,
+      "loss": 1.3223,
+      "step": 2710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.110668182373047,
+      "learning_rate": 4.9286016389109175e-05,
+      "loss": 1.6274,
+      "step": 2711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.990293025970459,
+      "learning_rate": 4.928575204863865e-05,
+      "loss": 1.4476,
+      "step": 2712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4308708906173706,
+      "learning_rate": 4.928548770816812e-05,
+      "loss": 1.472,
+      "step": 2713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.903432011604309,
+      "learning_rate": 4.9285223367697595e-05,
+      "loss": 1.4947,
+      "step": 2714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0009474754333496,
+      "learning_rate": 4.9284959027227076e-05,
+      "loss": 1.3274,
+      "step": 2715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5193769931793213,
+      "learning_rate": 4.928469468675654e-05,
+      "loss": 1.4444,
+      "step": 2716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4243419170379639,
+      "learning_rate": 4.928443034628602e-05,
+      "loss": 1.3634,
+      "step": 2717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4617148637771606,
+      "learning_rate": 4.928416600581549e-05,
+      "loss": 1.2577,
+      "step": 2718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5703520774841309,
+      "learning_rate": 4.928390166534497e-05,
+      "loss": 1.4556,
+      "step": 2719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7564822435379028,
+      "learning_rate": 4.9283637324874443e-05,
+      "loss": 1.3931,
+      "step": 2720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5954009294509888,
+      "learning_rate": 4.928337298440392e-05,
+      "loss": 1.4481,
+      "step": 2721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7772717475891113,
+      "learning_rate": 4.928310864393339e-05,
+      "loss": 1.3902,
+      "step": 2722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8575398921966553,
+      "learning_rate": 4.928284430346286e-05,
+      "loss": 1.3999,
+      "step": 2723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4367001056671143,
+      "learning_rate": 4.928257996299234e-05,
+      "loss": 1.3675,
+      "step": 2724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.459378957748413,
+      "learning_rate": 4.928231562252181e-05,
+      "loss": 2.5038,
+      "step": 2725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8390169143676758,
+      "learning_rate": 4.9282051282051285e-05,
+      "loss": 1.1387,
+      "step": 2726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.527518391609192,
+      "learning_rate": 4.928178694158076e-05,
+      "loss": 1.537,
+      "step": 2727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6860098838806152,
+      "learning_rate": 4.928152260111023e-05,
+      "loss": 1.4227,
+      "step": 2728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7111889123916626,
+      "learning_rate": 4.9281258260639706e-05,
+      "loss": 1.3938,
+      "step": 2729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4894232749938965,
+      "learning_rate": 4.9280993920169186e-05,
+      "loss": 1.4173,
+      "step": 2730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3538715839385986,
+      "learning_rate": 4.928072957969865e-05,
+      "loss": 2.0131,
+      "step": 2731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9447660446166992,
+      "learning_rate": 4.9280465239228126e-05,
+      "loss": 1.5895,
+      "step": 2732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6844242811203003,
+      "learning_rate": 4.92802008987576e-05,
+      "loss": 1.3779,
+      "step": 2733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2473037242889404,
+      "learning_rate": 4.9279936558287073e-05,
+      "loss": 1.7129,
+      "step": 2734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.124917984008789,
+      "learning_rate": 4.9279672217816554e-05,
+      "loss": 1.4031,
+      "step": 2735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7105717658996582,
+      "learning_rate": 4.927940787734602e-05,
+      "loss": 1.4819,
+      "step": 2736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6356290578842163,
+      "learning_rate": 4.92791435368755e-05,
+      "loss": 1.3946,
+      "step": 2737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4759626388549805,
+      "learning_rate": 4.927887919640497e-05,
+      "loss": 2.4185,
+      "step": 2738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1673524379730225,
+      "learning_rate": 4.927861485593445e-05,
+      "loss": 1.4262,
+      "step": 2739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8103241920471191,
+      "learning_rate": 4.927835051546392e-05,
+      "loss": 1.3518,
+      "step": 2740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1094696521759033,
+      "learning_rate": 4.9278086174993395e-05,
+      "loss": 1.5113,
+      "step": 2741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6180050373077393,
+      "learning_rate": 4.927782183452287e-05,
+      "loss": 1.375,
+      "step": 2742
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5468167066574097,
+      "learning_rate": 4.927755749405234e-05,
+      "loss": 1.3733,
+      "step": 2743
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.627955675125122,
+      "learning_rate": 4.9277293153581816e-05,
+      "loss": 1.454,
+      "step": 2744
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9188731908798218,
+      "learning_rate": 4.927702881311129e-05,
+      "loss": 1.5666,
+      "step": 2745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4699007272720337,
+      "learning_rate": 4.927676447264076e-05,
+      "loss": 1.4212,
+      "step": 2746
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.616644263267517,
+      "learning_rate": 4.927650013217024e-05,
+      "loss": 1.3441,
+      "step": 2747
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4245861768722534,
+      "learning_rate": 4.927623579169971e-05,
+      "loss": 1.3018,
+      "step": 2748
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5073254108428955,
+      "learning_rate": 4.9275971451229184e-05,
+      "loss": 1.4572,
+      "step": 2749
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.471834421157837,
+      "learning_rate": 4.9275707110758664e-05,
+      "loss": 1.2384,
+      "step": 2750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.53413987159729,
+      "learning_rate": 4.927544277028813e-05,
+      "loss": 1.4963,
+      "step": 2751
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.719245195388794,
+      "learning_rate": 4.927517842981761e-05,
+      "loss": 1.3792,
+      "step": 2752
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.363478422164917,
+      "learning_rate": 4.927491408934708e-05,
+      "loss": 1.3708,
+      "step": 2753
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3108142614364624,
+      "learning_rate": 4.927464974887656e-05,
+      "loss": 1.2405,
+      "step": 2754
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3397886753082275,
+      "learning_rate": 4.927438540840603e-05,
+      "loss": 1.1764,
+      "step": 2755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5279619693756104,
+      "learning_rate": 4.9274121067935506e-05,
+      "loss": 1.4124,
+      "step": 2756
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0691208839416504,
+      "learning_rate": 4.927385672746498e-05,
+      "loss": 1.5135,
+      "step": 2757
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.583957552909851,
+      "learning_rate": 4.927359238699445e-05,
+      "loss": 1.4016,
+      "step": 2758
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5294452905654907,
+      "learning_rate": 4.9273328046523926e-05,
+      "loss": 1.3285,
+      "step": 2759
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3332196474075317,
+      "learning_rate": 4.92730637060534e-05,
+      "loss": 1.3156,
+      "step": 2760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.808803081512451,
+      "learning_rate": 4.9272799365582873e-05,
+      "loss": 1.4987,
+      "step": 2761
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7851637601852417,
+      "learning_rate": 4.927253502511235e-05,
+      "loss": 1.3094,
+      "step": 2762
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6472556591033936,
+      "learning_rate": 4.927227068464182e-05,
+      "loss": 1.4247,
+      "step": 2763
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9602537155151367,
+      "learning_rate": 4.9272006344171294e-05,
+      "loss": 1.4758,
+      "step": 2764
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1466729640960693,
+      "learning_rate": 4.9271742003700774e-05,
+      "loss": 1.4457,
+      "step": 2765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.531456708908081,
+      "learning_rate": 4.927147766323024e-05,
+      "loss": 1.6034,
+      "step": 2766
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5536959171295166,
+      "learning_rate": 4.927121332275972e-05,
+      "loss": 1.3755,
+      "step": 2767
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.883734107017517,
+      "learning_rate": 4.927094898228919e-05,
+      "loss": 1.5138,
+      "step": 2768
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8666869401931763,
+      "learning_rate": 4.927068464181866e-05,
+      "loss": 1.5041,
+      "step": 2769
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6348004341125488,
+      "learning_rate": 4.927042030134814e-05,
+      "loss": 1.3306,
+      "step": 2770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4459971189498901,
+      "learning_rate": 4.927015596087761e-05,
+      "loss": 1.3707,
+      "step": 2771
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.814191460609436,
+      "learning_rate": 4.926989162040709e-05,
+      "loss": 1.4771,
+      "step": 2772
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.644896149635315,
+      "learning_rate": 4.9269627279936556e-05,
+      "loss": 1.3478,
+      "step": 2773
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5192862749099731,
+      "learning_rate": 4.926936293946604e-05,
+      "loss": 1.2071,
+      "step": 2774
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.62288498878479,
+      "learning_rate": 4.926909859899551e-05,
+      "loss": 1.3921,
+      "step": 2775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9504754543304443,
+      "learning_rate": 4.9268834258524984e-05,
+      "loss": 1.456,
+      "step": 2776
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.776131510734558,
+      "learning_rate": 4.926856991805446e-05,
+      "loss": 1.4387,
+      "step": 2777
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.145490884780884,
+      "learning_rate": 4.926830557758393e-05,
+      "loss": 1.3404,
+      "step": 2778
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3183660507202148,
+      "learning_rate": 4.9268041237113404e-05,
+      "loss": 1.2422,
+      "step": 2779
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3956469297409058,
+      "learning_rate": 4.926777689664288e-05,
+      "loss": 1.4096,
+      "step": 2780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7639714479446411,
+      "learning_rate": 4.926751255617235e-05,
+      "loss": 1.5102,
+      "step": 2781
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.863586664199829,
+      "learning_rate": 4.9267248215701825e-05,
+      "loss": 1.5436,
+      "step": 2782
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.566824197769165,
+      "learning_rate": 4.92669838752313e-05,
+      "loss": 1.5047,
+      "step": 2783
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.111873149871826,
+      "learning_rate": 4.926671953476077e-05,
+      "loss": 1.4761,
+      "step": 2784
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.831944465637207,
+      "learning_rate": 4.926645519429025e-05,
+      "loss": 1.4397,
+      "step": 2785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.164449691772461,
+      "learning_rate": 4.926619085381972e-05,
+      "loss": 1.3937,
+      "step": 2786
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7635068893432617,
+      "learning_rate": 4.92659265133492e-05,
+      "loss": 1.5119,
+      "step": 2787
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7318859100341797,
+      "learning_rate": 4.926566217287867e-05,
+      "loss": 1.4046,
+      "step": 2788
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8015538454055786,
+      "learning_rate": 4.926539783240815e-05,
+      "loss": 1.3752,
+      "step": 2789
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.203911304473877,
+      "learning_rate": 4.926513349193762e-05,
+      "loss": 1.4326,
+      "step": 2790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.518168568611145,
+      "learning_rate": 4.9264869151467094e-05,
+      "loss": 1.3245,
+      "step": 2791
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1006252765655518,
+      "learning_rate": 4.926460481099657e-05,
+      "loss": 1.5672,
+      "step": 2792
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.632002592086792,
+      "learning_rate": 4.926434047052604e-05,
+      "loss": 1.55,
+      "step": 2793
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9112212657928467,
+      "learning_rate": 4.9264076130055515e-05,
+      "loss": 1.5221,
+      "step": 2794
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.615892767906189,
+      "learning_rate": 4.926381178958499e-05,
+      "loss": 1.4967,
+      "step": 2795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8867878913879395,
+      "learning_rate": 4.926354744911446e-05,
+      "loss": 1.4468,
+      "step": 2796
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8549513816833496,
+      "learning_rate": 4.9263283108643936e-05,
+      "loss": 1.5063,
+      "step": 2797
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7279201745986938,
+      "learning_rate": 4.926301876817341e-05,
+      "loss": 1.4399,
+      "step": 2798
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.643559217453003,
+      "learning_rate": 4.926275442770288e-05,
+      "loss": 1.4412,
+      "step": 2799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5987133979797363,
+      "learning_rate": 4.926249008723236e-05,
+      "loss": 1.4191,
+      "step": 2800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5997995138168335,
+      "learning_rate": 4.926222574676183e-05,
+      "loss": 1.4937,
+      "step": 2801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.829963207244873,
+      "learning_rate": 4.926196140629131e-05,
+      "loss": 1.4206,
+      "step": 2802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.968198537826538,
+      "learning_rate": 4.926169706582078e-05,
+      "loss": 1.5695,
+      "step": 2803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9330799579620361,
+      "learning_rate": 4.926143272535025e-05,
+      "loss": 1.2968,
+      "step": 2804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3041179180145264,
+      "learning_rate": 4.926116838487973e-05,
+      "loss": 1.582,
+      "step": 2805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.569651484489441,
+      "learning_rate": 4.92609040444092e-05,
+      "loss": 1.3709,
+      "step": 2806
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.853461742401123,
+      "learning_rate": 4.926063970393868e-05,
+      "loss": 1.4664,
+      "step": 2807
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.812429666519165,
+      "learning_rate": 4.9260375363468145e-05,
+      "loss": 1.5148,
+      "step": 2808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5648832321166992,
+      "learning_rate": 4.9260111022997625e-05,
+      "loss": 1.8332,
+      "step": 2809
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2357027530670166,
+      "learning_rate": 4.92598466825271e-05,
+      "loss": 1.652,
+      "step": 2810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.714574933052063,
+      "learning_rate": 4.925958234205657e-05,
+      "loss": 1.2245,
+      "step": 2811
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5173749923706055,
+      "learning_rate": 4.9259318001586046e-05,
+      "loss": 1.5229,
+      "step": 2812
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5155857801437378,
+      "learning_rate": 4.925905366111552e-05,
+      "loss": 1.408,
+      "step": 2813
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7240917682647705,
+      "learning_rate": 4.925878932064499e-05,
+      "loss": 1.5408,
+      "step": 2814
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.980783462524414,
+      "learning_rate": 4.9258524980174467e-05,
+      "loss": 1.5149,
+      "step": 2815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6448007822036743,
+      "learning_rate": 4.925826063970394e-05,
+      "loss": 1.5662,
+      "step": 2816
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.504779815673828,
+      "learning_rate": 4.9257996299233414e-05,
+      "loss": 1.5171,
+      "step": 2817
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8992748260498047,
+      "learning_rate": 4.925773195876289e-05,
+      "loss": 1.345,
+      "step": 2818
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.895208477973938,
+      "learning_rate": 4.925746761829236e-05,
+      "loss": 1.4878,
+      "step": 2819
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.128929615020752,
+      "learning_rate": 4.925720327782184e-05,
+      "loss": 1.4077,
+      "step": 2820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.581985592842102,
+      "learning_rate": 4.925693893735131e-05,
+      "loss": 1.4036,
+      "step": 2821
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5514413118362427,
+      "learning_rate": 4.925667459688079e-05,
+      "loss": 1.3034,
+      "step": 2822
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8829073905944824,
+      "learning_rate": 4.9256410256410255e-05,
+      "loss": 1.3218,
+      "step": 2823
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2178752422332764,
+      "learning_rate": 4.9256145915939736e-05,
+      "loss": 1.6055,
+      "step": 2824
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9132566452026367,
+      "learning_rate": 4.925588157546921e-05,
+      "loss": 1.448,
+      "step": 2825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5989588499069214,
+      "learning_rate": 4.925561723499868e-05,
+      "loss": 1.4364,
+      "step": 2826
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.238628625869751,
+      "learning_rate": 4.9255352894528156e-05,
+      "loss": 1.5531,
+      "step": 2827
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.50469970703125,
+      "learning_rate": 4.925508855405763e-05,
+      "loss": 1.4921,
+      "step": 2828
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8935837745666504,
+      "learning_rate": 4.92548242135871e-05,
+      "loss": 1.3593,
+      "step": 2829
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7942612171173096,
+      "learning_rate": 4.925455987311658e-05,
+      "loss": 1.3202,
+      "step": 2830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8087096214294434,
+      "learning_rate": 4.925429553264605e-05,
+      "loss": 1.4768,
+      "step": 2831
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9597867727279663,
+      "learning_rate": 4.9254031192175524e-05,
+      "loss": 1.562,
+      "step": 2832
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.873789668083191,
+      "learning_rate": 4.9253766851705e-05,
+      "loss": 1.3892,
+      "step": 2833
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6122952699661255,
+      "learning_rate": 4.925350251123447e-05,
+      "loss": 1.5601,
+      "step": 2834
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0574951171875,
+      "learning_rate": 4.925323817076395e-05,
+      "loss": 1.5609,
+      "step": 2835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8174891471862793,
+      "learning_rate": 4.925297383029342e-05,
+      "loss": 1.5888,
+      "step": 2836
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.796237826347351,
+      "learning_rate": 4.92527094898229e-05,
+      "loss": 1.5497,
+      "step": 2837
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.681021809577942,
+      "learning_rate": 4.9252445149352365e-05,
+      "loss": 1.4103,
+      "step": 2838
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3721890449523926,
+      "learning_rate": 4.9252180808881846e-05,
+      "loss": 1.6554,
+      "step": 2839
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7418662309646606,
+      "learning_rate": 4.925191646841132e-05,
+      "loss": 2.0102,
+      "step": 2840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4495433568954468,
+      "learning_rate": 4.9251652127940786e-05,
+      "loss": 2.1211,
+      "step": 2841
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2741007804870605,
+      "learning_rate": 4.9251387787470267e-05,
+      "loss": 1.3302,
+      "step": 2842
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5786818265914917,
+      "learning_rate": 4.925112344699973e-05,
+      "loss": 1.368,
+      "step": 2843
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8689334392547607,
+      "learning_rate": 4.9250859106529214e-05,
+      "loss": 1.4405,
+      "step": 2844
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5736788511276245,
+      "learning_rate": 4.925059476605869e-05,
+      "loss": 1.3942,
+      "step": 2845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7176365852355957,
+      "learning_rate": 4.925033042558816e-05,
+      "loss": 1.3857,
+      "step": 2846
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5509557723999023,
+      "learning_rate": 4.9250066085117634e-05,
+      "loss": 1.4483,
+      "step": 2847
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9354842901229858,
+      "learning_rate": 4.924980174464711e-05,
+      "loss": 1.5645,
+      "step": 2848
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7870033979415894,
+      "learning_rate": 4.924953740417658e-05,
+      "loss": 1.4533,
+      "step": 2849
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6688764095306396,
+      "learning_rate": 4.9249273063706055e-05,
+      "loss": 1.3014,
+      "step": 2850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7349936962127686,
+      "learning_rate": 4.924900872323553e-05,
+      "loss": 1.5805,
+      "step": 2851
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5609962940216064,
+      "learning_rate": 4.9248744382765e-05,
+      "loss": 1.5121,
+      "step": 2852
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4994351863861084,
+      "learning_rate": 4.9248480042294476e-05,
+      "loss": 1.7247,
+      "step": 2853
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9496595859527588,
+      "learning_rate": 4.924821570182395e-05,
+      "loss": 1.5328,
+      "step": 2854
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.076174736022949,
+      "learning_rate": 4.924795136135343e-05,
+      "loss": 1.4124,
+      "step": 2855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6105163097381592,
+      "learning_rate": 4.9247687020882897e-05,
+      "loss": 1.4427,
+      "step": 2856
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.575422763824463,
+      "learning_rate": 4.924742268041238e-05,
+      "loss": 2.1088,
+      "step": 2857
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8424683809280396,
+      "learning_rate": 4.9247158339941844e-05,
+      "loss": 1.4234,
+      "step": 2858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2485994100570679,
+      "learning_rate": 4.9246893999471324e-05,
+      "loss": 1.2529,
+      "step": 2859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5964714288711548,
+      "learning_rate": 4.92466296590008e-05,
+      "loss": 1.8717,
+      "step": 2860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.921461343765259,
+      "learning_rate": 4.924636531853027e-05,
+      "loss": 1.6105,
+      "step": 2861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8925305604934692,
+      "learning_rate": 4.9246100978059745e-05,
+      "loss": 1.4862,
+      "step": 2862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7731595039367676,
+      "learning_rate": 4.924583663758922e-05,
+      "loss": 1.3981,
+      "step": 2863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6928892135620117,
+      "learning_rate": 4.924557229711869e-05,
+      "loss": 1.4124,
+      "step": 2864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7464566230773926,
+      "learning_rate": 4.9245307956648165e-05,
+      "loss": 1.2313,
+      "step": 2865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.683985710144043,
+      "learning_rate": 4.924504361617764e-05,
+      "loss": 1.147,
+      "step": 2866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8656883239746094,
+      "learning_rate": 4.924477927570711e-05,
+      "loss": 1.3841,
+      "step": 2867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.211930274963379,
+      "learning_rate": 4.9244514935236586e-05,
+      "loss": 1.5244,
+      "step": 2868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3886908292770386,
+      "learning_rate": 4.924425059476606e-05,
+      "loss": 1.3861,
+      "step": 2869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6073412895202637,
+      "learning_rate": 4.924398625429554e-05,
+      "loss": 1.4307,
+      "step": 2870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9059364795684814,
+      "learning_rate": 4.924372191382501e-05,
+      "loss": 1.4333,
+      "step": 2871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6473846435546875,
+      "learning_rate": 4.924345757335449e-05,
+      "loss": 1.409,
+      "step": 2872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9787110090255737,
+      "learning_rate": 4.9243193232883954e-05,
+      "loss": 1.4491,
+      "step": 2873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6183462142944336,
+      "learning_rate": 4.9242928892413434e-05,
+      "loss": 1.5702,
+      "step": 2874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7224377393722534,
+      "learning_rate": 4.924266455194291e-05,
+      "loss": 1.3718,
+      "step": 2875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8962537050247192,
+      "learning_rate": 4.924240021147238e-05,
+      "loss": 1.4397,
+      "step": 2876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.377892255783081,
+      "learning_rate": 4.9242135871001855e-05,
+      "loss": 1.5313,
+      "step": 2877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.478628158569336,
+      "learning_rate": 4.924187153053132e-05,
+      "loss": 1.3393,
+      "step": 2878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7981359958648682,
+      "learning_rate": 4.92416071900608e-05,
+      "loss": 1.5501,
+      "step": 2879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6880958080291748,
+      "learning_rate": 4.9241342849590276e-05,
+      "loss": 1.3619,
+      "step": 2880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7978668212890625,
+      "learning_rate": 4.924107850911975e-05,
+      "loss": 1.4482,
+      "step": 2881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.842490553855896,
+      "learning_rate": 4.924081416864922e-05,
+      "loss": 1.4573,
+      "step": 2882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5252575874328613,
+      "learning_rate": 4.9240549828178697e-05,
+      "loss": 1.432,
+      "step": 2883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5570218563079834,
+      "learning_rate": 4.924028548770817e-05,
+      "loss": 1.2724,
+      "step": 2884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7761777639389038,
+      "learning_rate": 4.924002114723765e-05,
+      "loss": 1.4914,
+      "step": 2885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5392897129058838,
+      "learning_rate": 4.923975680676712e-05,
+      "loss": 1.4761,
+      "step": 2886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1829352378845215,
+      "learning_rate": 4.923949246629659e-05,
+      "loss": 1.4836,
+      "step": 2887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8341717720031738,
+      "learning_rate": 4.9239228125826064e-05,
+      "loss": 1.4024,
+      "step": 2888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8544707298278809,
+      "learning_rate": 4.923896378535554e-05,
+      "loss": 1.4669,
+      "step": 2889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.654718279838562,
+      "learning_rate": 4.923869944488502e-05,
+      "loss": 1.5866,
+      "step": 2890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.131880044937134,
+      "learning_rate": 4.9238435104414485e-05,
+      "loss": 1.4975,
+      "step": 2891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7268786430358887,
+      "learning_rate": 4.9238170763943965e-05,
+      "loss": 1.8372,
+      "step": 2892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5116640329360962,
+      "learning_rate": 4.923790642347343e-05,
+      "loss": 1.3329,
+      "step": 2893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8628979921340942,
+      "learning_rate": 4.923764208300291e-05,
+      "loss": 1.4053,
+      "step": 2894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.593846321105957,
+      "learning_rate": 4.9237377742532386e-05,
+      "loss": 1.4555,
+      "step": 2895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.699266791343689,
+      "learning_rate": 4.923711340206186e-05,
+      "loss": 1.4473,
+      "step": 2896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5363372564315796,
+      "learning_rate": 4.923684906159133e-05,
+      "loss": 1.3829,
+      "step": 2897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.646601676940918,
+      "learning_rate": 4.923658472112081e-05,
+      "loss": 1.3017,
+      "step": 2898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5237030982971191,
+      "learning_rate": 4.923632038065028e-05,
+      "loss": 1.4924,
+      "step": 2899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6375207901000977,
+      "learning_rate": 4.9236056040179754e-05,
+      "loss": 1.2951,
+      "step": 2900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5321402549743652,
+      "learning_rate": 4.923579169970923e-05,
+      "loss": 1.3824,
+      "step": 2901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3855071067810059,
+      "learning_rate": 4.92355273592387e-05,
+      "loss": 1.3511,
+      "step": 2902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1732535362243652,
+      "learning_rate": 4.9235263018768175e-05,
+      "loss": 1.5216,
+      "step": 2903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8809077739715576,
+      "learning_rate": 4.923499867829765e-05,
+      "loss": 1.7824,
+      "step": 2904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5217012166976929,
+      "learning_rate": 4.923473433782713e-05,
+      "loss": 1.3162,
+      "step": 2905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4964635372161865,
+      "learning_rate": 4.9234469997356595e-05,
+      "loss": 1.4245,
+      "step": 2906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7991535663604736,
+      "learning_rate": 4.9234205656886076e-05,
+      "loss": 1.6049,
+      "step": 2907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.866365671157837,
+      "learning_rate": 4.923394131641554e-05,
+      "loss": 1.4845,
+      "step": 2908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4478416442871094,
+      "learning_rate": 4.923367697594502e-05,
+      "loss": 1.5099,
+      "step": 2909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.976857304573059,
+      "learning_rate": 4.9233412635474496e-05,
+      "loss": 1.4112,
+      "step": 2910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0588934421539307,
+      "learning_rate": 4.923314829500397e-05,
+      "loss": 1.5224,
+      "step": 2911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.534363031387329,
+      "learning_rate": 4.9232883954533444e-05,
+      "loss": 1.3973,
+      "step": 2912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0758743286132812,
+      "learning_rate": 4.923261961406291e-05,
+      "loss": 1.3612,
+      "step": 2913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1561310291290283,
+      "learning_rate": 4.923235527359239e-05,
+      "loss": 1.5107,
+      "step": 2914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.928382396697998,
+      "learning_rate": 4.9232090933121864e-05,
+      "loss": 1.4637,
+      "step": 2915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.032317876815796,
+      "learning_rate": 4.923182659265134e-05,
+      "loss": 1.5061,
+      "step": 2916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0053820610046387,
+      "learning_rate": 4.923156225218081e-05,
+      "loss": 1.291,
+      "step": 2917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7419453859329224,
+      "learning_rate": 4.9231297911710285e-05,
+      "loss": 1.3986,
+      "step": 2918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8612408638000488,
+      "learning_rate": 4.923103357123976e-05,
+      "loss": 1.3636,
+      "step": 2919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5189950466156006,
+      "learning_rate": 4.923076923076924e-05,
+      "loss": 1.3917,
+      "step": 2920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7555196285247803,
+      "learning_rate": 4.9230504890298706e-05,
+      "loss": 1.4699,
+      "step": 2921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1041078567504883,
+      "learning_rate": 4.923024054982818e-05,
+      "loss": 1.673,
+      "step": 2922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.302535891532898,
+      "learning_rate": 4.922997620935765e-05,
+      "loss": 1.2113,
+      "step": 2923
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7077146768569946,
+      "learning_rate": 4.9229711868887126e-05,
+      "loss": 1.3652,
+      "step": 2924
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3319337368011475,
+      "learning_rate": 4.922944752841661e-05,
+      "loss": 1.3081,
+      "step": 2925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8819637298583984,
+      "learning_rate": 4.9229183187946074e-05,
+      "loss": 1.5752,
+      "step": 2926
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8818998336791992,
+      "learning_rate": 4.9228918847475554e-05,
+      "loss": 1.3836,
+      "step": 2927
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5064128637313843,
+      "learning_rate": 4.922865450700502e-05,
+      "loss": 2.5654,
+      "step": 2928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.521054983139038,
+      "learning_rate": 4.92283901665345e-05,
+      "loss": 1.3386,
+      "step": 2929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.690869688987732,
+      "learning_rate": 4.9228125826063975e-05,
+      "loss": 1.3999,
+      "step": 2930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.864231586456299,
+      "learning_rate": 4.922786148559345e-05,
+      "loss": 1.487,
+      "step": 2931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0483710765838623,
+      "learning_rate": 4.922759714512292e-05,
+      "loss": 1.6718,
+      "step": 2932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0830979347229004,
+      "learning_rate": 4.9227332804652395e-05,
+      "loss": 1.4885,
+      "step": 2933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9728875160217285,
+      "learning_rate": 4.922706846418187e-05,
+      "loss": 2.5842,
+      "step": 2934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.980332374572754,
+      "learning_rate": 4.922680412371134e-05,
+      "loss": 1.5157,
+      "step": 2935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5280462503433228,
+      "learning_rate": 4.9226539783240816e-05,
+      "loss": 1.3483,
+      "step": 2936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6889257431030273,
+      "learning_rate": 4.922627544277029e-05,
+      "loss": 1.5934,
+      "step": 2937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0816936492919922,
+      "learning_rate": 4.922601110229976e-05,
+      "loss": 1.3166,
+      "step": 2938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.60575270652771,
+      "learning_rate": 4.922574676182924e-05,
+      "loss": 1.4574,
+      "step": 2939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5258870124816895,
+      "learning_rate": 4.922548242135872e-05,
+      "loss": 1.4689,
+      "step": 2940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.84442138671875,
+      "learning_rate": 4.9225218080888184e-05,
+      "loss": 1.5299,
+      "step": 2941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4847745895385742,
+      "learning_rate": 4.9224953740417664e-05,
+      "loss": 1.4207,
+      "step": 2942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4377012252807617,
+      "learning_rate": 4.922468939994713e-05,
+      "loss": 1.5104,
+      "step": 2943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4100525379180908,
+      "learning_rate": 4.922442505947661e-05,
+      "loss": 1.2904,
+      "step": 2944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6369329690933228,
+      "learning_rate": 4.9224160719006085e-05,
+      "loss": 1.403,
+      "step": 2945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.759497046470642,
+      "learning_rate": 4.922389637853556e-05,
+      "loss": 1.5126,
+      "step": 2946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4238636493682861,
+      "learning_rate": 4.922363203806503e-05,
+      "loss": 1.4183,
+      "step": 2947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9247169494628906,
+      "learning_rate": 4.9223367697594506e-05,
+      "loss": 1.6158,
+      "step": 2948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.496205449104309,
+      "learning_rate": 4.922310335712398e-05,
+      "loss": 1.3142,
+      "step": 2949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7374366521835327,
+      "learning_rate": 4.922283901665345e-05,
+      "loss": 1.4452,
+      "step": 2950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6415561437606812,
+      "learning_rate": 4.9222574676182926e-05,
+      "loss": 1.542,
+      "step": 2951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4476869106292725,
+      "learning_rate": 4.92223103357124e-05,
+      "loss": 1.585,
+      "step": 2952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5890296697616577,
+      "learning_rate": 4.9222045995241874e-05,
+      "loss": 1.3573,
+      "step": 2953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6343231201171875,
+      "learning_rate": 4.922178165477135e-05,
+      "loss": 1.3814,
+      "step": 2954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5826010704040527,
+      "learning_rate": 4.922151731430083e-05,
+      "loss": 1.4162,
+      "step": 2955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6332570314407349,
+      "learning_rate": 4.9221252973830294e-05,
+      "loss": 1.5204,
+      "step": 2956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6370124816894531,
+      "learning_rate": 4.9220988633359775e-05,
+      "loss": 1.3964,
+      "step": 2957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.07089900970459,
+      "learning_rate": 4.922072429288924e-05,
+      "loss": 1.4468,
+      "step": 2958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4706412553787231,
+      "learning_rate": 4.9220459952418715e-05,
+      "loss": 1.5232,
+      "step": 2959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8595151901245117,
+      "learning_rate": 4.9220195611948195e-05,
+      "loss": 1.5181,
+      "step": 2960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.821195363998413,
+      "learning_rate": 4.921993127147766e-05,
+      "loss": 1.6017,
+      "step": 2961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.607391119003296,
+      "learning_rate": 4.921966693100714e-05,
+      "loss": 1.4334,
+      "step": 2962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0832769870758057,
+      "learning_rate": 4.921940259053661e-05,
+      "loss": 1.4937,
+      "step": 2963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6252517700195312,
+      "learning_rate": 4.921913825006609e-05,
+      "loss": 1.4146,
+      "step": 2964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9847251176834106,
+      "learning_rate": 4.921887390959556e-05,
+      "loss": 1.3907,
+      "step": 2965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3725016117095947,
+      "learning_rate": 4.921860956912504e-05,
+      "loss": 1.453,
+      "step": 2966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7191359996795654,
+      "learning_rate": 4.921834522865451e-05,
+      "loss": 1.2966,
+      "step": 2967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.261263132095337,
+      "learning_rate": 4.9218080888183984e-05,
+      "loss": 1.5096,
+      "step": 2968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7134041786193848,
+      "learning_rate": 4.921781654771346e-05,
+      "loss": 1.3669,
+      "step": 2969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.528936743736267,
+      "learning_rate": 4.921755220724293e-05,
+      "loss": 1.4527,
+      "step": 2970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7915802001953125,
+      "learning_rate": 4.9217287866772405e-05,
+      "loss": 1.3101,
+      "step": 2971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7619177103042603,
+      "learning_rate": 4.921702352630188e-05,
+      "loss": 1.5207,
+      "step": 2972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0681631565093994,
+      "learning_rate": 4.921675918583135e-05,
+      "loss": 1.388,
+      "step": 2973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7841145992279053,
+      "learning_rate": 4.9216494845360825e-05,
+      "loss": 1.4215,
+      "step": 2974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8098480701446533,
+      "learning_rate": 4.92162305048903e-05,
+      "loss": 1.4928,
+      "step": 2975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7207143306732178,
+      "learning_rate": 4.921596616441977e-05,
+      "loss": 1.3583,
+      "step": 2976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.678417205810547,
+      "learning_rate": 4.921570182394925e-05,
+      "loss": 1.5444,
+      "step": 2977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4058802127838135,
+      "learning_rate": 4.921543748347872e-05,
+      "loss": 1.488,
+      "step": 2978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9101197719573975,
+      "learning_rate": 4.92151731430082e-05,
+      "loss": 1.3045,
+      "step": 2979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3600165843963623,
+      "learning_rate": 4.921490880253767e-05,
+      "loss": 1.3247,
+      "step": 2980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3812836408615112,
+      "learning_rate": 4.921464446206715e-05,
+      "loss": 1.3976,
+      "step": 2981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8383501768112183,
+      "learning_rate": 4.921438012159662e-05,
+      "loss": 1.4661,
+      "step": 2982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5993808507919312,
+      "learning_rate": 4.9214115781126094e-05,
+      "loss": 1.393,
+      "step": 2983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4990359544754028,
+      "learning_rate": 4.921385144065557e-05,
+      "loss": 1.4708,
+      "step": 2984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8648897409439087,
+      "learning_rate": 4.9213587100185035e-05,
+      "loss": 1.2342,
+      "step": 2985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.707838773727417,
+      "learning_rate": 4.9213322759714515e-05,
+      "loss": 1.4832,
+      "step": 2986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.46136474609375,
+      "learning_rate": 4.921305841924399e-05,
+      "loss": 1.4577,
+      "step": 2987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6125576496124268,
+      "learning_rate": 4.921279407877346e-05,
+      "loss": 1.4457,
+      "step": 2988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5153992176055908,
+      "learning_rate": 4.9212529738302936e-05,
+      "loss": 1.4424,
+      "step": 2989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9607782363891602,
+      "learning_rate": 4.921226539783241e-05,
+      "loss": 1.5315,
+      "step": 2990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4200150966644287,
+      "learning_rate": 4.921200105736188e-05,
+      "loss": 1.2673,
+      "step": 2991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0157933235168457,
+      "learning_rate": 4.921173671689136e-05,
+      "loss": 1.5571,
+      "step": 2992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2880414724349976,
+      "learning_rate": 4.921147237642083e-05,
+      "loss": 1.4398,
+      "step": 2993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2476415634155273,
+      "learning_rate": 4.921120803595031e-05,
+      "loss": 1.5228,
+      "step": 2994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7628201246261597,
+      "learning_rate": 4.921094369547978e-05,
+      "loss": 1.4458,
+      "step": 2995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.305140972137451,
+      "learning_rate": 4.921067935500925e-05,
+      "loss": 1.5512,
+      "step": 2996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5672755241394043,
+      "learning_rate": 4.921041501453873e-05,
+      "loss": 1.398,
+      "step": 2997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.44049870967865,
+      "learning_rate": 4.92101506740682e-05,
+      "loss": 1.4674,
+      "step": 2998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3238886594772339,
+      "learning_rate": 4.920988633359768e-05,
+      "loss": 1.3224,
+      "step": 2999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5795272588729858,
+      "learning_rate": 4.9209621993127145e-05,
+      "loss": 1.2792,
+      "step": 3000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.089759111404419,
+      "learning_rate": 4.9209357652656625e-05,
+      "loss": 1.357,
+      "step": 3001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5320755243301392,
+      "learning_rate": 4.92090933121861e-05,
+      "loss": 1.2916,
+      "step": 3002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7346091270446777,
+      "learning_rate": 4.920882897171557e-05,
+      "loss": 1.3046,
+      "step": 3003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5792076587677002,
+      "learning_rate": 4.9208564631245046e-05,
+      "loss": 1.3762,
+      "step": 3004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.496769428253174,
+      "learning_rate": 4.920830029077452e-05,
+      "loss": 1.5321,
+      "step": 3005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5207581520080566,
+      "learning_rate": 4.920803595030399e-05,
+      "loss": 1.6905,
+      "step": 3006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3846497535705566,
+      "learning_rate": 4.920777160983347e-05,
+      "loss": 1.392,
+      "step": 3007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6408969163894653,
+      "learning_rate": 4.920750726936294e-05,
+      "loss": 1.4142,
+      "step": 3008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.273242712020874,
+      "learning_rate": 4.9207242928892414e-05,
+      "loss": 1.529,
+      "step": 3009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5079829692840576,
+      "learning_rate": 4.920697858842189e-05,
+      "loss": 1.318,
+      "step": 3010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5117218494415283,
+      "learning_rate": 4.920671424795136e-05,
+      "loss": 1.3871,
+      "step": 3011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.033987283706665,
+      "learning_rate": 4.920644990748084e-05,
+      "loss": 1.5242,
+      "step": 3012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.675106406211853,
+      "learning_rate": 4.920618556701031e-05,
+      "loss": 1.4444,
+      "step": 3013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.010232925415039,
+      "learning_rate": 4.920592122653979e-05,
+      "loss": 1.4972,
+      "step": 3014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.453675389289856,
+      "learning_rate": 4.9205656886069255e-05,
+      "loss": 2.6245,
+      "step": 3015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6367837190628052,
+      "learning_rate": 4.9205392545598736e-05,
+      "loss": 1.2721,
+      "step": 3016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7838866710662842,
+      "learning_rate": 4.920512820512821e-05,
+      "loss": 1.4939,
+      "step": 3017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9595108032226562,
+      "learning_rate": 4.920486386465768e-05,
+      "loss": 1.3985,
+      "step": 3018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7957842350006104,
+      "learning_rate": 4.9204599524187156e-05,
+      "loss": 1.4145,
+      "step": 3019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4364662170410156,
+      "learning_rate": 4.920433518371663e-05,
+      "loss": 1.2514,
+      "step": 3020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3533984422683716,
+      "learning_rate": 4.9204070843246104e-05,
+      "loss": 1.9618,
+      "step": 3021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9187395572662354,
+      "learning_rate": 4.920380650277558e-05,
+      "loss": 1.5161,
+      "step": 3022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6019504070281982,
+      "learning_rate": 4.920354216230505e-05,
+      "loss": 1.4143,
+      "step": 3023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.042738676071167,
+      "learning_rate": 4.9203277821834524e-05,
+      "loss": 1.3363,
+      "step": 3024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.729324460029602,
+      "learning_rate": 4.9203013481364e-05,
+      "loss": 1.4551,
+      "step": 3025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.974212646484375,
+      "learning_rate": 4.920274914089347e-05,
+      "loss": 1.6399,
+      "step": 3026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1369552612304688,
+      "learning_rate": 4.920248480042295e-05,
+      "loss": 1.2915,
+      "step": 3027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5995044708251953,
+      "learning_rate": 4.920222045995242e-05,
+      "loss": 1.4765,
+      "step": 3028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5540961027145386,
+      "learning_rate": 4.92019561194819e-05,
+      "loss": 1.494,
+      "step": 3029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7059065103530884,
+      "learning_rate": 4.9201691779011366e-05,
+      "loss": 1.9104,
+      "step": 3030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6065447330474854,
+      "learning_rate": 4.920142743854084e-05,
+      "loss": 1.573,
+      "step": 3031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7574020624160767,
+      "learning_rate": 4.920116309807032e-05,
+      "loss": 1.4205,
+      "step": 3032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8597532510757446,
+      "learning_rate": 4.9200898757599786e-05,
+      "loss": 2.1139,
+      "step": 3033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7847745418548584,
+      "learning_rate": 4.920063441712927e-05,
+      "loss": 1.4489,
+      "step": 3034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.433520555496216,
+      "learning_rate": 4.9200370076658734e-05,
+      "loss": 1.4516,
+      "step": 3035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7747039794921875,
+      "learning_rate": 4.9200105736188214e-05,
+      "loss": 1.359,
+      "step": 3036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5744597911834717,
+      "learning_rate": 4.919984139571769e-05,
+      "loss": 1.4006,
+      "step": 3037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0844690799713135,
+      "learning_rate": 4.919957705524716e-05,
+      "loss": 1.5994,
+      "step": 3038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5507302284240723,
+      "learning_rate": 4.9199312714776635e-05,
+      "loss": 1.4459,
+      "step": 3039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6989787817001343,
+      "learning_rate": 4.919904837430611e-05,
+      "loss": 1.5112,
+      "step": 3040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.641439437866211,
+      "learning_rate": 4.919878403383558e-05,
+      "loss": 1.3828,
+      "step": 3041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4782737493515015,
+      "learning_rate": 4.9198519693365055e-05,
+      "loss": 1.1874,
+      "step": 3042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4847456216812134,
+      "learning_rate": 4.919825535289453e-05,
+      "loss": 1.3525,
+      "step": 3043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.64609432220459,
+      "learning_rate": 4.9197991012424e-05,
+      "loss": 1.6276,
+      "step": 3044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2897419929504395,
+      "learning_rate": 4.9197726671953476e-05,
+      "loss": 1.2589,
+      "step": 3045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1006827354431152,
+      "learning_rate": 4.919746233148295e-05,
+      "loss": 1.4684,
+      "step": 3046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7540751695632935,
+      "learning_rate": 4.919719799101243e-05,
+      "loss": 1.4785,
+      "step": 3047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5938103199005127,
+      "learning_rate": 4.91969336505419e-05,
+      "loss": 1.4441,
+      "step": 3048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8222485780715942,
+      "learning_rate": 4.919666931007138e-05,
+      "loss": 1.4134,
+      "step": 3049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7987921237945557,
+      "learning_rate": 4.9196404969600844e-05,
+      "loss": 1.4858,
+      "step": 3050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4751837253570557,
+      "learning_rate": 4.9196140629130324e-05,
+      "loss": 1.3936,
+      "step": 3051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7848037481307983,
+      "learning_rate": 4.91958762886598e-05,
+      "loss": 1.437,
+      "step": 3052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8393665552139282,
+      "learning_rate": 4.919561194818927e-05,
+      "loss": 1.5652,
+      "step": 3053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6371169090270996,
+      "learning_rate": 4.9195347607718745e-05,
+      "loss": 1.3959,
+      "step": 3054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.414589285850525,
+      "learning_rate": 4.919508326724822e-05,
+      "loss": 1.3184,
+      "step": 3055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4782696962356567,
+      "learning_rate": 4.919481892677769e-05,
+      "loss": 1.2461,
+      "step": 3056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5758823156356812,
+      "learning_rate": 4.9194554586307166e-05,
+      "loss": 1.5132,
+      "step": 3057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8425129652023315,
+      "learning_rate": 4.919429024583664e-05,
+      "loss": 1.4644,
+      "step": 3058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2837893962860107,
+      "learning_rate": 4.919402590536611e-05,
+      "loss": 1.4837,
+      "step": 3059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5396177768707275,
+      "learning_rate": 4.9193761564895586e-05,
+      "loss": 1.4491,
+      "step": 3060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.985414743423462,
+      "learning_rate": 4.919349722442506e-05,
+      "loss": 1.3967,
+      "step": 3061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.441499948501587,
+      "learning_rate": 4.919323288395454e-05,
+      "loss": 1.4939,
+      "step": 3062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5319738388061523,
+      "learning_rate": 4.919296854348401e-05,
+      "loss": 1.3331,
+      "step": 3063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8294085264205933,
+      "learning_rate": 4.919270420301349e-05,
+      "loss": 1.4081,
+      "step": 3064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4292387962341309,
+      "learning_rate": 4.9192439862542954e-05,
+      "loss": 2.4462,
+      "step": 3065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8540416955947876,
+      "learning_rate": 4.9192175522072435e-05,
+      "loss": 1.4882,
+      "step": 3066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4023574590682983,
+      "learning_rate": 4.919191118160191e-05,
+      "loss": 1.3275,
+      "step": 3067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.627519965171814,
+      "learning_rate": 4.9191646841131375e-05,
+      "loss": 1.3458,
+      "step": 3068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.775039553642273,
+      "learning_rate": 4.9191382500660855e-05,
+      "loss": 1.4655,
+      "step": 3069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4481301307678223,
+      "learning_rate": 4.919111816019032e-05,
+      "loss": 1.3715,
+      "step": 3070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.988422155380249,
+      "learning_rate": 4.91908538197198e-05,
+      "loss": 1.541,
+      "step": 3071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.854104995727539,
+      "learning_rate": 4.9190589479249276e-05,
+      "loss": 1.5726,
+      "step": 3072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8904412984848022,
+      "learning_rate": 4.919032513877875e-05,
+      "loss": 1.3899,
+      "step": 3073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.446821928024292,
+      "learning_rate": 4.919006079830822e-05,
+      "loss": 1.3805,
+      "step": 3074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.665772795677185,
+      "learning_rate": 4.91897964578377e-05,
+      "loss": 1.3307,
+      "step": 3075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4049513339996338,
+      "learning_rate": 4.918953211736717e-05,
+      "loss": 2.2176,
+      "step": 3076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.58913254737854,
+      "learning_rate": 4.9189267776896644e-05,
+      "loss": 1.355,
+      "step": 3077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4786268472671509,
+      "learning_rate": 4.918900343642612e-05,
+      "loss": 1.298,
+      "step": 3078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7910226583480835,
+      "learning_rate": 4.918873909595559e-05,
+      "loss": 1.3816,
+      "step": 3079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3325908184051514,
+      "learning_rate": 4.9188474755485065e-05,
+      "loss": 1.6234,
+      "step": 3080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.921899676322937,
+      "learning_rate": 4.918821041501454e-05,
+      "loss": 1.4315,
+      "step": 3081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6207568645477295,
+      "learning_rate": 4.918794607454402e-05,
+      "loss": 1.5649,
+      "step": 3082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4165045022964478,
+      "learning_rate": 4.9187681734073485e-05,
+      "loss": 1.304,
+      "step": 3083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2059028148651123,
+      "learning_rate": 4.9187417393602966e-05,
+      "loss": 1.4583,
+      "step": 3084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.859028697013855,
+      "learning_rate": 4.918715305313243e-05,
+      "loss": 1.1952,
+      "step": 3085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5493011474609375,
+      "learning_rate": 4.918688871266191e-05,
+      "loss": 1.6472,
+      "step": 3086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6733733415603638,
+      "learning_rate": 4.9186624372191386e-05,
+      "loss": 1.325,
+      "step": 3087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.053752899169922,
+      "learning_rate": 4.918636003172086e-05,
+      "loss": 1.4637,
+      "step": 3088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6098722219467163,
+      "learning_rate": 4.9186095691250333e-05,
+      "loss": 1.3255,
+      "step": 3089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5996736288070679,
+      "learning_rate": 4.918583135077981e-05,
+      "loss": 1.3036,
+      "step": 3090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.610783338546753,
+      "learning_rate": 4.918556701030928e-05,
+      "loss": 1.4582,
+      "step": 3091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4929218292236328,
+      "learning_rate": 4.9185302669838754e-05,
+      "loss": 1.3808,
+      "step": 3092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6663594245910645,
+      "learning_rate": 4.918503832936823e-05,
+      "loss": 1.3321,
+      "step": 3093
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.461343288421631,
+      "learning_rate": 4.91847739888977e-05,
+      "loss": 1.5309,
+      "step": 3094
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5419731140136719,
+      "learning_rate": 4.9184509648427175e-05,
+      "loss": 1.5072,
+      "step": 3095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5480642318725586,
+      "learning_rate": 4.918424530795665e-05,
+      "loss": 1.3161,
+      "step": 3096
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7944599390029907,
+      "learning_rate": 4.918398096748613e-05,
+      "loss": 1.6448,
+      "step": 3097
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8102102279663086,
+      "learning_rate": 4.9183716627015596e-05,
+      "loss": 1.5578,
+      "step": 3098
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.358106017112732,
+      "learning_rate": 4.9183452286545076e-05,
+      "loss": 1.347,
+      "step": 3099
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.494465708732605,
+      "learning_rate": 4.918318794607454e-05,
+      "loss": 1.4676,
+      "step": 3100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.286500930786133,
+      "learning_rate": 4.918292360560402e-05,
+      "loss": 1.4969,
+      "step": 3101
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6576491594314575,
+      "learning_rate": 4.91826592651335e-05,
+      "loss": 1.4966,
+      "step": 3102
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8604682683944702,
+      "learning_rate": 4.9182394924662963e-05,
+      "loss": 1.2206,
+      "step": 3103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4149994850158691,
+      "learning_rate": 4.9182130584192444e-05,
+      "loss": 1.4763,
+      "step": 3104
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6765987873077393,
+      "learning_rate": 4.918186624372191e-05,
+      "loss": 1.4412,
+      "step": 3105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.837925672531128,
+      "learning_rate": 4.918160190325139e-05,
+      "loss": 1.4371,
+      "step": 3106
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5168683528900146,
+      "learning_rate": 4.9181337562780864e-05,
+      "loss": 1.2916,
+      "step": 3107
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3397135734558105,
+      "learning_rate": 4.918107322231034e-05,
+      "loss": 1.1829,
+      "step": 3108
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0079944133758545,
+      "learning_rate": 4.918080888183981e-05,
+      "loss": 1.5645,
+      "step": 3109
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5605775117874146,
+      "learning_rate": 4.9180544541369285e-05,
+      "loss": 1.2965,
+      "step": 3110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9087415933609009,
+      "learning_rate": 4.918028020089876e-05,
+      "loss": 1.4205,
+      "step": 3111
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.654736876487732,
+      "learning_rate": 4.918001586042823e-05,
+      "loss": 1.4506,
+      "step": 3112
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5775067806243896,
+      "learning_rate": 4.9179751519957706e-05,
+      "loss": 1.4285,
+      "step": 3113
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7045869827270508,
+      "learning_rate": 4.917948717948718e-05,
+      "loss": 1.5807,
+      "step": 3114
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8057575225830078,
+      "learning_rate": 4.917922283901665e-05,
+      "loss": 1.5599,
+      "step": 3115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.559609293937683,
+      "learning_rate": 4.917895849854613e-05,
+      "loss": 1.3841,
+      "step": 3116
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9628225564956665,
+      "learning_rate": 4.917869415807561e-05,
+      "loss": 1.4291,
+      "step": 3117
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.943281650543213,
+      "learning_rate": 4.9178429817605074e-05,
+      "loss": 1.5554,
+      "step": 3118
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.005122423171997,
+      "learning_rate": 4.9178165477134554e-05,
+      "loss": 1.5352,
+      "step": 3119
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6521461009979248,
+      "learning_rate": 4.917790113666402e-05,
+      "loss": 1.5281,
+      "step": 3120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7121652364730835,
+      "learning_rate": 4.91776367961935e-05,
+      "loss": 1.331,
+      "step": 3121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7799127101898193,
+      "learning_rate": 4.9177372455722975e-05,
+      "loss": 1.3615,
+      "step": 3122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2810232639312744,
+      "learning_rate": 4.917710811525245e-05,
+      "loss": 1.3836,
+      "step": 3123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1367878913879395,
+      "learning_rate": 4.917684377478192e-05,
+      "loss": 1.4289,
+      "step": 3124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2366809844970703,
+      "learning_rate": 4.9176579434311396e-05,
+      "loss": 1.6157,
+      "step": 3125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9745864868164062,
+      "learning_rate": 4.917631509384087e-05,
+      "loss": 2.4219,
+      "step": 3126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5877797603607178,
+      "learning_rate": 4.917605075337034e-05,
+      "loss": 1.3015,
+      "step": 3127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9861087799072266,
+      "learning_rate": 4.9175786412899816e-05,
+      "loss": 1.4263,
+      "step": 3128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3675413131713867,
+      "learning_rate": 4.917552207242929e-05,
+      "loss": 1.4243,
+      "step": 3129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2848172187805176,
+      "learning_rate": 4.9175257731958763e-05,
+      "loss": 1.4193,
+      "step": 3130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9663152694702148,
+      "learning_rate": 4.917499339148824e-05,
+      "loss": 1.5314,
+      "step": 3131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6578519344329834,
+      "learning_rate": 4.917472905101772e-05,
+      "loss": 1.4736,
+      "step": 3132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7002614736557007,
+      "learning_rate": 4.9174464710547184e-05,
+      "loss": 1.4479,
+      "step": 3133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3127691745758057,
+      "learning_rate": 4.9174200370076664e-05,
+      "loss": 1.1934,
+      "step": 3134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.051499128341675,
+      "learning_rate": 4.917393602960613e-05,
+      "loss": 1.4013,
+      "step": 3135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2878507375717163,
+      "learning_rate": 4.917367168913561e-05,
+      "loss": 1.5645,
+      "step": 3136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.978918194770813,
+      "learning_rate": 4.9173407348665085e-05,
+      "loss": 1.5171,
+      "step": 3137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0446484088897705,
+      "learning_rate": 4.917314300819456e-05,
+      "loss": 1.5009,
+      "step": 3138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.385550856590271,
+      "learning_rate": 4.917287866772403e-05,
+      "loss": 1.8334,
+      "step": 3139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4885871410369873,
+      "learning_rate": 4.91726143272535e-05,
+      "loss": 1.3538,
+      "step": 3140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.576137065887451,
+      "learning_rate": 4.917234998678298e-05,
+      "loss": 1.387,
+      "step": 3141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.771655797958374,
+      "learning_rate": 4.917208564631245e-05,
+      "loss": 1.5047,
+      "step": 3142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5768259763717651,
+      "learning_rate": 4.9171821305841927e-05,
+      "loss": 1.3553,
+      "step": 3143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5313313007354736,
+      "learning_rate": 4.91715569653714e-05,
+      "loss": 1.4476,
+      "step": 3144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.018477201461792,
+      "learning_rate": 4.9171292624900874e-05,
+      "loss": 1.4258,
+      "step": 3145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3637398481369019,
+      "learning_rate": 4.917102828443035e-05,
+      "loss": 1.2985,
+      "step": 3146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6170244216918945,
+      "learning_rate": 4.917076394395983e-05,
+      "loss": 1.4628,
+      "step": 3147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4316370487213135,
+      "learning_rate": 4.9170499603489294e-05,
+      "loss": 1.4521,
+      "step": 3148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4649858474731445,
+      "learning_rate": 4.917023526301877e-05,
+      "loss": 1.2943,
+      "step": 3149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.456979513168335,
+      "learning_rate": 4.916997092254824e-05,
+      "loss": 1.4116,
+      "step": 3150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.473214864730835,
+      "learning_rate": 4.9169706582077715e-05,
+      "loss": 1.3738,
+      "step": 3151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9568746089935303,
+      "learning_rate": 4.9169442241607196e-05,
+      "loss": 1.5378,
+      "step": 3152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4262017011642456,
+      "learning_rate": 4.916917790113666e-05,
+      "loss": 1.3369,
+      "step": 3153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6769468784332275,
+      "learning_rate": 4.916891356066614e-05,
+      "loss": 1.3172,
+      "step": 3154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7380681037902832,
+      "learning_rate": 4.916864922019561e-05,
+      "loss": 1.4434,
+      "step": 3155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.941290020942688,
+      "learning_rate": 4.916838487972509e-05,
+      "loss": 1.4678,
+      "step": 3156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8797295093536377,
+      "learning_rate": 4.916812053925456e-05,
+      "loss": 1.429,
+      "step": 3157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7141978740692139,
+      "learning_rate": 4.916785619878404e-05,
+      "loss": 1.2786,
+      "step": 3158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7318333387374878,
+      "learning_rate": 4.916759185831351e-05,
+      "loss": 1.4537,
+      "step": 3159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7043036222457886,
+      "learning_rate": 4.9167327517842984e-05,
+      "loss": 1.4291,
+      "step": 3160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9726872444152832,
+      "learning_rate": 4.916706317737246e-05,
+      "loss": 1.4783,
+      "step": 3161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7167021036148071,
+      "learning_rate": 4.916679883690193e-05,
+      "loss": 1.3839,
+      "step": 3162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8159431219100952,
+      "learning_rate": 4.9166534496431405e-05,
+      "loss": 1.3246,
+      "step": 3163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7099392414093018,
+      "learning_rate": 4.916627015596088e-05,
+      "loss": 1.4865,
+      "step": 3164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.344770669937134,
+      "learning_rate": 4.916600581549035e-05,
+      "loss": 1.5332,
+      "step": 3165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.674837589263916,
+      "learning_rate": 4.9165741475019826e-05,
+      "loss": 1.2521,
+      "step": 3166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5366284847259521,
+      "learning_rate": 4.9165477134549306e-05,
+      "loss": 1.2007,
+      "step": 3167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7548062801361084,
+      "learning_rate": 4.916521279407877e-05,
+      "loss": 1.6482,
+      "step": 3168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9423490762710571,
+      "learning_rate": 4.916494845360825e-05,
+      "loss": 1.4251,
+      "step": 3169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.74578058719635,
+      "learning_rate": 4.916468411313772e-05,
+      "loss": 1.4675,
+      "step": 3170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4903807640075684,
+      "learning_rate": 4.91644197726672e-05,
+      "loss": 1.3002,
+      "step": 3171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.366441488265991,
+      "learning_rate": 4.9164155432196674e-05,
+      "loss": 1.6345,
+      "step": 3172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5568233728408813,
+      "learning_rate": 4.916389109172615e-05,
+      "loss": 1.3599,
+      "step": 3173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9148705005645752,
+      "learning_rate": 4.916362675125562e-05,
+      "loss": 1.6138,
+      "step": 3174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9023367166519165,
+      "learning_rate": 4.916336241078509e-05,
+      "loss": 1.2509,
+      "step": 3175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4455785751342773,
+      "learning_rate": 4.916309807031457e-05,
+      "loss": 1.3003,
+      "step": 3176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1777501106262207,
+      "learning_rate": 4.916283372984404e-05,
+      "loss": 1.6943,
+      "step": 3177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3868451118469238,
+      "learning_rate": 4.9162569389373515e-05,
+      "loss": 1.4116,
+      "step": 3178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7963231801986694,
+      "learning_rate": 4.916230504890299e-05,
+      "loss": 1.4534,
+      "step": 3179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5629572868347168,
+      "learning_rate": 4.916204070843246e-05,
+      "loss": 1.4177,
+      "step": 3180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5716992616653442,
+      "learning_rate": 4.9161776367961936e-05,
+      "loss": 1.4018,
+      "step": 3181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9528138637542725,
+      "learning_rate": 4.9161512027491416e-05,
+      "loss": 1.5802,
+      "step": 3182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8374444246292114,
+      "learning_rate": 4.916124768702088e-05,
+      "loss": 1.4432,
+      "step": 3183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6015357971191406,
+      "learning_rate": 4.916098334655036e-05,
+      "loss": 1.2951,
+      "step": 3184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5129586458206177,
+      "learning_rate": 4.916071900607983e-05,
+      "loss": 1.2759,
+      "step": 3185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9867244958877563,
+      "learning_rate": 4.9160454665609304e-05,
+      "loss": 1.6006,
+      "step": 3186
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.668318748474121,
+      "learning_rate": 4.9160190325138784e-05,
+      "loss": 1.4885,
+      "step": 3187
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4788950681686401,
+      "learning_rate": 4.915992598466825e-05,
+      "loss": 1.1956,
+      "step": 3188
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.188814163208008,
+      "learning_rate": 4.915966164419773e-05,
+      "loss": 2.6911,
+      "step": 3189
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5451022386550903,
+      "learning_rate": 4.91593973037272e-05,
+      "loss": 1.3301,
+      "step": 3190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7188342809677124,
+      "learning_rate": 4.915913296325668e-05,
+      "loss": 1.438,
+      "step": 3191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5695425271987915,
+      "learning_rate": 4.915886862278615e-05,
+      "loss": 1.3958,
+      "step": 3192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8171900510787964,
+      "learning_rate": 4.9158604282315625e-05,
+      "loss": 1.3775,
+      "step": 3193
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7408462762832642,
+      "learning_rate": 4.91583399418451e-05,
+      "loss": 1.3518,
+      "step": 3194
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8519452810287476,
+      "learning_rate": 4.915807560137457e-05,
+      "loss": 1.3573,
+      "step": 3195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7982823848724365,
+      "learning_rate": 4.9157811260904046e-05,
+      "loss": 1.8748,
+      "step": 3196
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9416509866714478,
+      "learning_rate": 4.915754692043352e-05,
+      "loss": 1.4467,
+      "step": 3197
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5978753566741943,
+      "learning_rate": 4.915728257996299e-05,
+      "loss": 1.4477,
+      "step": 3198
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8075602054595947,
+      "learning_rate": 4.915701823949247e-05,
+      "loss": 1.3996,
+      "step": 3199
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.781821370124817,
+      "learning_rate": 4.915675389902194e-05,
+      "loss": 1.4293,
+      "step": 3200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.417263388633728,
+      "learning_rate": 4.9156489558551414e-05,
+      "loss": 1.4546,
+      "step": 3201
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8427517414093018,
+      "learning_rate": 4.9156225218080894e-05,
+      "loss": 1.3817,
+      "step": 3202
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0794944763183594,
+      "learning_rate": 4.915596087761036e-05,
+      "loss": 1.8636,
+      "step": 3203
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6165825128555298,
+      "learning_rate": 4.915569653713984e-05,
+      "loss": 1.295,
+      "step": 3204
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0111610889434814,
+      "learning_rate": 4.915543219666931e-05,
+      "loss": 1.4584,
+      "step": 3205
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5218688249588013,
+      "learning_rate": 4.915516785619879e-05,
+      "loss": 1.4104,
+      "step": 3206
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4677079916000366,
+      "learning_rate": 4.915490351572826e-05,
+      "loss": 1.3632,
+      "step": 3207
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8637326955795288,
+      "learning_rate": 4.9154639175257736e-05,
+      "loss": 1.5789,
+      "step": 3208
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9598585367202759,
+      "learning_rate": 4.915437483478721e-05,
+      "loss": 1.4706,
+      "step": 3209
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9554498195648193,
+      "learning_rate": 4.915411049431668e-05,
+      "loss": 2.1457,
+      "step": 3210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8122951984405518,
+      "learning_rate": 4.9153846153846157e-05,
+      "loss": 1.2801,
+      "step": 3211
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6782225370407104,
+      "learning_rate": 4.915358181337563e-05,
+      "loss": 1.505,
+      "step": 3212
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.445233702659607,
+      "learning_rate": 4.9153317472905104e-05,
+      "loss": 1.1468,
+      "step": 3213
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.559535264968872,
+      "learning_rate": 4.915305313243458e-05,
+      "loss": 1.4419,
+      "step": 3214
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9001859426498413,
+      "learning_rate": 4.915278879196405e-05,
+      "loss": 1.8339,
+      "step": 3215
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0280444622039795,
+      "learning_rate": 4.9152524451493524e-05,
+      "loss": 1.4474,
+      "step": 3216
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5471458435058594,
+      "learning_rate": 4.9152260111023005e-05,
+      "loss": 1.372,
+      "step": 3217
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6280531883239746,
+      "learning_rate": 4.915199577055247e-05,
+      "loss": 1.3404,
+      "step": 3218
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.033513069152832,
+      "learning_rate": 4.915173143008195e-05,
+      "loss": 1.5312,
+      "step": 3219
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3777180910110474,
+      "learning_rate": 4.915146708961142e-05,
+      "loss": 1.2408,
+      "step": 3220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2524950504302979,
+      "learning_rate": 4.915120274914089e-05,
+      "loss": 1.1601,
+      "step": 3221
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.430241584777832,
+      "learning_rate": 4.915093840867037e-05,
+      "loss": 1.3493,
+      "step": 3222
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3283438682556152,
+      "learning_rate": 4.915067406819984e-05,
+      "loss": 1.3566,
+      "step": 3223
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1475613117218018,
+      "learning_rate": 4.915040972772932e-05,
+      "loss": 1.5838,
+      "step": 3224
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2579848766326904,
+      "learning_rate": 4.9150145387258787e-05,
+      "loss": 1.4681,
+      "step": 3225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7580938339233398,
+      "learning_rate": 4.914988104678827e-05,
+      "loss": 1.5065,
+      "step": 3226
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9097986221313477,
+      "learning_rate": 4.914961670631774e-05,
+      "loss": 1.4449,
+      "step": 3227
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9730992317199707,
+      "learning_rate": 4.9149352365847214e-05,
+      "loss": 1.5306,
+      "step": 3228
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3147411346435547,
+      "learning_rate": 4.914908802537669e-05,
+      "loss": 1.3442,
+      "step": 3229
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.046342134475708,
+      "learning_rate": 4.914882368490616e-05,
+      "loss": 1.4215,
+      "step": 3230
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.150402307510376,
+      "learning_rate": 4.9148559344435635e-05,
+      "loss": 1.5591,
+      "step": 3231
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.69166100025177,
+      "learning_rate": 4.914829500396511e-05,
+      "loss": 1.2752,
+      "step": 3232
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9357454776763916,
+      "learning_rate": 4.914803066349458e-05,
+      "loss": 1.5015,
+      "step": 3233
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5221753120422363,
+      "learning_rate": 4.9147766323024055e-05,
+      "loss": 1.4693,
+      "step": 3234
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6403300762176514,
+      "learning_rate": 4.914750198255353e-05,
+      "loss": 1.4364,
+      "step": 3235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9763957262039185,
+      "learning_rate": 4.9147237642083e-05,
+      "loss": 1.4597,
+      "step": 3236
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5771266222000122,
+      "learning_rate": 4.914697330161248e-05,
+      "loss": 1.4276,
+      "step": 3237
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4220530986785889,
+      "learning_rate": 4.914670896114195e-05,
+      "loss": 1.2851,
+      "step": 3238
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6313509941101074,
+      "learning_rate": 4.914644462067143e-05,
+      "loss": 1.3768,
+      "step": 3239
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6803056001663208,
+      "learning_rate": 4.91461802802009e-05,
+      "loss": 1.2906,
+      "step": 3240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6018099784851074,
+      "learning_rate": 4.914591593973038e-05,
+      "loss": 1.5088,
+      "step": 3241
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.452317714691162,
+      "learning_rate": 4.914565159925985e-05,
+      "loss": 1.3087,
+      "step": 3242
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3377715349197388,
+      "learning_rate": 4.9145387258789324e-05,
+      "loss": 1.3263,
+      "step": 3243
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5025744438171387,
+      "learning_rate": 4.91451229183188e-05,
+      "loss": 1.4806,
+      "step": 3244
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6748523712158203,
+      "learning_rate": 4.914485857784827e-05,
+      "loss": 1.4722,
+      "step": 3245
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6339279413223267,
+      "learning_rate": 4.9144594237377745e-05,
+      "loss": 1.4578,
+      "step": 3246
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9369170665740967,
+      "learning_rate": 4.914432989690722e-05,
+      "loss": 1.5716,
+      "step": 3247
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8765220642089844,
+      "learning_rate": 4.914406555643669e-05,
+      "loss": 1.5422,
+      "step": 3248
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8948625326156616,
+      "learning_rate": 4.9143801215966166e-05,
+      "loss": 1.5808,
+      "step": 3249
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2421224117279053,
+      "learning_rate": 4.914353687549564e-05,
+      "loss": 1.5753,
+      "step": 3250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8688305616378784,
+      "learning_rate": 4.914327253502511e-05,
+      "loss": 1.3606,
+      "step": 3251
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7641068696975708,
+      "learning_rate": 4.914300819455459e-05,
+      "loss": 1.421,
+      "step": 3252
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6477622985839844,
+      "learning_rate": 4.914274385408406e-05,
+      "loss": 1.6047,
+      "step": 3253
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3930234909057617,
+      "learning_rate": 4.914247951361354e-05,
+      "loss": 1.4031,
+      "step": 3254
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0105185508728027,
+      "learning_rate": 4.914221517314301e-05,
+      "loss": 1.4634,
+      "step": 3255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4137369394302368,
+      "learning_rate": 4.914195083267249e-05,
+      "loss": 1.4375,
+      "step": 3256
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.874062180519104,
+      "learning_rate": 4.914168649220196e-05,
+      "loss": 1.4598,
+      "step": 3257
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6034963130950928,
+      "learning_rate": 4.914142215173143e-05,
+      "loss": 1.43,
+      "step": 3258
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5453732013702393,
+      "learning_rate": 4.914115781126091e-05,
+      "loss": 1.3603,
+      "step": 3259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9742413759231567,
+      "learning_rate": 4.9140893470790375e-05,
+      "loss": 1.4251,
+      "step": 3260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4300150871276855,
+      "learning_rate": 4.9140629130319855e-05,
+      "loss": 1.2054,
+      "step": 3261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6239519119262695,
+      "learning_rate": 4.914036478984933e-05,
+      "loss": 1.5684,
+      "step": 3262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5476958751678467,
+      "learning_rate": 4.91401004493788e-05,
+      "loss": 1.498,
+      "step": 3263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9114902019500732,
+      "learning_rate": 4.9139836108908276e-05,
+      "loss": 1.3673,
+      "step": 3264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6381433010101318,
+      "learning_rate": 4.913957176843775e-05,
+      "loss": 1.3409,
+      "step": 3265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.085568428039551,
+      "learning_rate": 4.913930742796722e-05,
+      "loss": 1.3747,
+      "step": 3266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7972794771194458,
+      "learning_rate": 4.91390430874967e-05,
+      "loss": 1.5517,
+      "step": 3267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0132970809936523,
+      "learning_rate": 4.913877874702617e-05,
+      "loss": 1.6086,
+      "step": 3268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9938790798187256,
+      "learning_rate": 4.9138514406555644e-05,
+      "loss": 1.4693,
+      "step": 3269
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4254201650619507,
+      "learning_rate": 4.913825006608512e-05,
+      "loss": 1.4861,
+      "step": 3270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5605796575546265,
+      "learning_rate": 4.913798572561459e-05,
+      "loss": 1.3397,
+      "step": 3271
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.386965274810791,
+      "learning_rate": 4.913772138514407e-05,
+      "loss": 2.0918,
+      "step": 3272
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4804741144180298,
+      "learning_rate": 4.913745704467354e-05,
+      "loss": 1.3606,
+      "step": 3273
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7376203536987305,
+      "learning_rate": 4.913719270420302e-05,
+      "loss": 1.4615,
+      "step": 3274
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4816069602966309,
+      "learning_rate": 4.9136928363732485e-05,
+      "loss": 1.3451,
+      "step": 3275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9809516668319702,
+      "learning_rate": 4.9136664023261966e-05,
+      "loss": 1.5013,
+      "step": 3276
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6114975214004517,
+      "learning_rate": 4.913639968279144e-05,
+      "loss": 1.4653,
+      "step": 3277
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8667525053024292,
+      "learning_rate": 4.913613534232091e-05,
+      "loss": 1.5619,
+      "step": 3278
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0224692821502686,
+      "learning_rate": 4.9135871001850386e-05,
+      "loss": 1.3354,
+      "step": 3279
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4908206462860107,
+      "learning_rate": 4.913560666137986e-05,
+      "loss": 1.3426,
+      "step": 3280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7923479080200195,
+      "learning_rate": 4.9135342320909334e-05,
+      "loss": 1.4208,
+      "step": 3281
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6596369743347168,
+      "learning_rate": 4.913507798043881e-05,
+      "loss": 2.6451,
+      "step": 3282
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4157495498657227,
+      "learning_rate": 4.913481363996828e-05,
+      "loss": 1.5691,
+      "step": 3283
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6427720785140991,
+      "learning_rate": 4.9134549299497754e-05,
+      "loss": 1.5187,
+      "step": 3284
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.772212266921997,
+      "learning_rate": 4.913428495902723e-05,
+      "loss": 2.5895,
+      "step": 3285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6518152952194214,
+      "learning_rate": 4.91340206185567e-05,
+      "loss": 1.5512,
+      "step": 3286
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.018505096435547,
+      "learning_rate": 4.913375627808618e-05,
+      "loss": 1.4483,
+      "step": 3287
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7428553104400635,
+      "learning_rate": 4.913349193761565e-05,
+      "loss": 1.2441,
+      "step": 3288
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.517168641090393,
+      "learning_rate": 4.913322759714513e-05,
+      "loss": 1.3653,
+      "step": 3289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1170144081115723,
+      "learning_rate": 4.9132963256674596e-05,
+      "loss": 1.5788,
+      "step": 3290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0210423469543457,
+      "learning_rate": 4.9132698916204076e-05,
+      "loss": 1.5539,
+      "step": 3291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.514060378074646,
+      "learning_rate": 4.913243457573355e-05,
+      "loss": 1.5276,
+      "step": 3292
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6963766813278198,
+      "learning_rate": 4.9132170235263016e-05,
+      "loss": 1.4277,
+      "step": 3293
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.941443681716919,
+      "learning_rate": 4.91319058947925e-05,
+      "loss": 1.4061,
+      "step": 3294
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.821285367012024,
+      "learning_rate": 4.9131641554321964e-05,
+      "loss": 1.8593,
+      "step": 3295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6727685928344727,
+      "learning_rate": 4.9131377213851444e-05,
+      "loss": 1.3187,
+      "step": 3296
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7487508058547974,
+      "learning_rate": 4.913111287338092e-05,
+      "loss": 1.4926,
+      "step": 3297
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5821820497512817,
+      "learning_rate": 4.913084853291039e-05,
+      "loss": 1.3515,
+      "step": 3298
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6093902587890625,
+      "learning_rate": 4.9130584192439865e-05,
+      "loss": 1.6639,
+      "step": 3299
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8657870292663574,
+      "learning_rate": 4.913031985196934e-05,
+      "loss": 1.5166,
+      "step": 3300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.051011085510254,
+      "learning_rate": 4.913005551149881e-05,
+      "loss": 1.5448,
+      "step": 3301
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3900740146636963,
+      "learning_rate": 4.912979117102829e-05,
+      "loss": 1.5058,
+      "step": 3302
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8487460613250732,
+      "learning_rate": 4.912952683055776e-05,
+      "loss": 1.3601,
+      "step": 3303
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.531849980354309,
+      "learning_rate": 4.912926249008723e-05,
+      "loss": 1.3974,
+      "step": 3304
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0542173385620117,
+      "learning_rate": 4.9128998149616706e-05,
+      "loss": 1.4888,
+      "step": 3305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9519404172897339,
+      "learning_rate": 4.912873380914618e-05,
+      "loss": 1.5803,
+      "step": 3306
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8459330797195435,
+      "learning_rate": 4.912846946867566e-05,
+      "loss": 1.5071,
+      "step": 3307
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2820870876312256,
+      "learning_rate": 4.912820512820513e-05,
+      "loss": 1.3343,
+      "step": 3308
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7305225133895874,
+      "learning_rate": 4.912794078773461e-05,
+      "loss": 1.4489,
+      "step": 3309
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5974102020263672,
+      "learning_rate": 4.9127676447264074e-05,
+      "loss": 1.7633,
+      "step": 3310
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9457635879516602,
+      "learning_rate": 4.9127412106793554e-05,
+      "loss": 1.5166,
+      "step": 3311
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7672890424728394,
+      "learning_rate": 4.912714776632303e-05,
+      "loss": 1.523,
+      "step": 3312
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6644186973571777,
+      "learning_rate": 4.91268834258525e-05,
+      "loss": 1.3983,
+      "step": 3313
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8776382207870483,
+      "learning_rate": 4.9126619085381975e-05,
+      "loss": 1.4928,
+      "step": 3314
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4255433082580566,
+      "learning_rate": 4.912635474491145e-05,
+      "loss": 1.57,
+      "step": 3315
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0964415073394775,
+      "learning_rate": 4.912609040444092e-05,
+      "loss": 1.503,
+      "step": 3316
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9718246459960938,
+      "learning_rate": 4.9125826063970396e-05,
+      "loss": 1.4968,
+      "step": 3317
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8197768926620483,
+      "learning_rate": 4.912556172349987e-05,
+      "loss": 1.2667,
+      "step": 3318
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.608903169631958,
+      "learning_rate": 4.912529738302934e-05,
+      "loss": 1.3835,
+      "step": 3319
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8504215478897095,
+      "learning_rate": 4.9125033042558816e-05,
+      "loss": 1.5561,
+      "step": 3320
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.17585563659668,
+      "learning_rate": 4.912476870208829e-05,
+      "loss": 2.3456,
+      "step": 3321
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4983736276626587,
+      "learning_rate": 4.912450436161777e-05,
+      "loss": 1.3569,
+      "step": 3322
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3892003297805786,
+      "learning_rate": 4.912424002114724e-05,
+      "loss": 2.2795,
+      "step": 3323
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4427677392959595,
+      "learning_rate": 4.912397568067672e-05,
+      "loss": 1.2548,
+      "step": 3324
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0288097858428955,
+      "learning_rate": 4.9123711340206184e-05,
+      "loss": 1.4725,
+      "step": 3325
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2859582901000977,
+      "learning_rate": 4.9123446999735665e-05,
+      "loss": 1.4899,
+      "step": 3326
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.753248929977417,
+      "learning_rate": 4.912318265926514e-05,
+      "loss": 1.4449,
+      "step": 3327
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.692211389541626,
+      "learning_rate": 4.912291831879461e-05,
+      "loss": 1.3081,
+      "step": 3328
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6943691968917847,
+      "learning_rate": 4.9122653978324085e-05,
+      "loss": 1.4069,
+      "step": 3329
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.175459146499634,
+      "learning_rate": 4.912238963785355e-05,
+      "loss": 1.5036,
+      "step": 3330
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3864885568618774,
+      "learning_rate": 4.912212529738303e-05,
+      "loss": 1.2988,
+      "step": 3331
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.739288330078125,
+      "learning_rate": 4.9121860956912506e-05,
+      "loss": 1.3442,
+      "step": 3332
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7261301279067993,
+      "learning_rate": 4.912159661644198e-05,
+      "loss": 1.4702,
+      "step": 3333
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8006924390792847,
+      "learning_rate": 4.912133227597145e-05,
+      "loss": 1.3158,
+      "step": 3334
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7816940546035767,
+      "learning_rate": 4.912106793550093e-05,
+      "loss": 1.4676,
+      "step": 3335
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.744446039199829,
+      "learning_rate": 4.91208035950304e-05,
+      "loss": 1.3516,
+      "step": 3336
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9010294675827026,
+      "learning_rate": 4.912053925455988e-05,
+      "loss": 1.2929,
+      "step": 3337
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9991878271102905,
+      "learning_rate": 4.912027491408935e-05,
+      "loss": 1.3631,
+      "step": 3338
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.820182204246521,
+      "learning_rate": 4.912001057361882e-05,
+      "loss": 1.3864,
+      "step": 3339
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1435487270355225,
+      "learning_rate": 4.9119746233148295e-05,
+      "loss": 1.5193,
+      "step": 3340
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0497119426727295,
+      "learning_rate": 4.911948189267777e-05,
+      "loss": 1.4431,
+      "step": 3341
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.741330623626709,
+      "learning_rate": 4.911921755220725e-05,
+      "loss": 1.3422,
+      "step": 3342
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0753214359283447,
+      "learning_rate": 4.9118953211736715e-05,
+      "loss": 1.352,
+      "step": 3343
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.785772681236267,
+      "learning_rate": 4.9118688871266196e-05,
+      "loss": 1.3389,
+      "step": 3344
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9653631448745728,
+      "learning_rate": 4.911842453079566e-05,
+      "loss": 1.5667,
+      "step": 3345
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8040616512298584,
+      "learning_rate": 4.911816019032514e-05,
+      "loss": 1.318,
+      "step": 3346
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1690993309020996,
+      "learning_rate": 4.9117895849854616e-05,
+      "loss": 1.7402,
+      "step": 3347
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9117470979690552,
+      "learning_rate": 4.911763150938409e-05,
+      "loss": 1.4718,
+      "step": 3348
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.286086320877075,
+      "learning_rate": 4.9117367168913564e-05,
+      "loss": 1.2575,
+      "step": 3349
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.508338451385498,
+      "learning_rate": 4.911710282844304e-05,
+      "loss": 1.4499,
+      "step": 3350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8018500804901123,
+      "learning_rate": 4.911683848797251e-05,
+      "loss": 1.4364,
+      "step": 3351
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8854808807373047,
+      "learning_rate": 4.9116574147501984e-05,
+      "loss": 1.3304,
+      "step": 3352
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8490352630615234,
+      "learning_rate": 4.911630980703146e-05,
+      "loss": 1.4191,
+      "step": 3353
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8064792156219482,
+      "learning_rate": 4.911604546656093e-05,
+      "loss": 1.5115,
+      "step": 3354
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8084521293640137,
+      "learning_rate": 4.9115781126090405e-05,
+      "loss": 1.4144,
+      "step": 3355
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9618881940841675,
+      "learning_rate": 4.911551678561988e-05,
+      "loss": 1.5212,
+      "step": 3356
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.620943546295166,
+      "learning_rate": 4.911525244514936e-05,
+      "loss": 1.6249,
+      "step": 3357
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6223875284194946,
+      "learning_rate": 4.9114988104678826e-05,
+      "loss": 1.5337,
+      "step": 3358
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.508668899536133,
+      "learning_rate": 4.9114723764208306e-05,
+      "loss": 2.2445,
+      "step": 3359
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5073299407958984,
+      "learning_rate": 4.911445942373777e-05,
+      "loss": 1.1518,
+      "step": 3360
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8690693378448486,
+      "learning_rate": 4.911419508326725e-05,
+      "loss": 1.3432,
+      "step": 3361
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6399550437927246,
+      "learning_rate": 4.911393074279673e-05,
+      "loss": 1.5339,
+      "step": 3362
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.537663221359253,
+      "learning_rate": 4.91136664023262e-05,
+      "loss": 1.5514,
+      "step": 3363
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6143083572387695,
+      "learning_rate": 4.9113402061855674e-05,
+      "loss": 1.346,
+      "step": 3364
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.744997978210449,
+      "learning_rate": 4.911313772138514e-05,
+      "loss": 1.4376,
+      "step": 3365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.060727596282959,
+      "learning_rate": 4.911287338091462e-05,
+      "loss": 1.2686,
+      "step": 3366
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5930134057998657,
+      "learning_rate": 4.9112609040444095e-05,
+      "loss": 1.3659,
+      "step": 3367
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9661556482315063,
+      "learning_rate": 4.911234469997357e-05,
+      "loss": 1.4607,
+      "step": 3368
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8033181428909302,
+      "learning_rate": 4.911208035950304e-05,
+      "loss": 1.404,
+      "step": 3369
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4879226684570312,
+      "learning_rate": 4.9111816019032515e-05,
+      "loss": 1.4905,
+      "step": 3370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7752208709716797,
+      "learning_rate": 4.911155167856199e-05,
+      "loss": 1.3535,
+      "step": 3371
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.749942421913147,
+      "learning_rate": 4.911128733809147e-05,
+      "loss": 1.3553,
+      "step": 3372
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8571712970733643,
+      "learning_rate": 4.9111022997620936e-05,
+      "loss": 1.4476,
+      "step": 3373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1151182651519775,
+      "learning_rate": 4.9110758657150416e-05,
+      "loss": 1.4889,
+      "step": 3374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3742940425872803,
+      "learning_rate": 4.911049431667988e-05,
+      "loss": 1.6296,
+      "step": 3375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.63302743434906,
+      "learning_rate": 4.911022997620936e-05,
+      "loss": 1.2311,
+      "step": 3376
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.439751625061035,
+      "learning_rate": 4.910996563573884e-05,
+      "loss": 1.4897,
+      "step": 3377
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6533417701721191,
+      "learning_rate": 4.9109701295268304e-05,
+      "loss": 1.3572,
+      "step": 3378
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9105700254440308,
+      "learning_rate": 4.9109436954797784e-05,
+      "loss": 1.4641,
+      "step": 3379
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4800490140914917,
+      "learning_rate": 4.910917261432725e-05,
+      "loss": 1.5359,
+      "step": 3380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.655508041381836,
+      "learning_rate": 4.910890827385673e-05,
+      "loss": 1.4828,
+      "step": 3381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.103015422821045,
+      "learning_rate": 4.9108643933386205e-05,
+      "loss": 1.5405,
+      "step": 3382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4189863204956055,
+      "learning_rate": 4.910837959291568e-05,
+      "loss": 1.5161,
+      "step": 3383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8561630249023438,
+      "learning_rate": 4.910811525244515e-05,
+      "loss": 1.4373,
+      "step": 3384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6996207237243652,
+      "learning_rate": 4.9107850911974626e-05,
+      "loss": 1.4979,
+      "step": 3385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.762204647064209,
+      "learning_rate": 4.91075865715041e-05,
+      "loss": 1.3562,
+      "step": 3386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1198348999023438,
+      "learning_rate": 4.910732223103357e-05,
+      "loss": 1.5635,
+      "step": 3387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9049865007400513,
+      "learning_rate": 4.9107057890563046e-05,
+      "loss": 1.5366,
+      "step": 3388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7694823741912842,
+      "learning_rate": 4.910679355009252e-05,
+      "loss": 1.5338,
+      "step": 3389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5152195692062378,
+      "learning_rate": 4.9106529209621993e-05,
+      "loss": 1.3178,
+      "step": 3390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7790446281433105,
+      "learning_rate": 4.910626486915147e-05,
+      "loss": 1.1451,
+      "step": 3391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0174238681793213,
+      "learning_rate": 4.910600052868095e-05,
+      "loss": 1.5114,
+      "step": 3392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.842821717262268,
+      "learning_rate": 4.9105736188210414e-05,
+      "loss": 1.4132,
+      "step": 3393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8093602657318115,
+      "learning_rate": 4.9105471847739895e-05,
+      "loss": 1.5054,
+      "step": 3394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.433239221572876,
+      "learning_rate": 4.910520750726936e-05,
+      "loss": 1.5217,
+      "step": 3395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.740270733833313,
+      "learning_rate": 4.910494316679884e-05,
+      "loss": 1.3871,
+      "step": 3396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9734365940093994,
+      "learning_rate": 4.9104678826328315e-05,
+      "loss": 1.4528,
+      "step": 3397
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5967379808425903,
+      "learning_rate": 4.910441448585779e-05,
+      "loss": 1.4624,
+      "step": 3398
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4421205520629883,
+      "learning_rate": 4.910415014538726e-05,
+      "loss": 2.2562,
+      "step": 3399
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7572219371795654,
+      "learning_rate": 4.9103885804916736e-05,
+      "loss": 1.5918,
+      "step": 3400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4213823080062866,
+      "learning_rate": 4.910362146444621e-05,
+      "loss": 1.4189,
+      "step": 3401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5591394901275635,
+      "learning_rate": 4.910335712397568e-05,
+      "loss": 2.4479,
+      "step": 3402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5292620658874512,
+      "learning_rate": 4.910309278350516e-05,
+      "loss": 1.4878,
+      "step": 3403
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.006037950515747,
+      "learning_rate": 4.910282844303463e-05,
+      "loss": 1.5719,
+      "step": 3404
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4869487285614014,
+      "learning_rate": 4.9102564102564104e-05,
+      "loss": 1.2766,
+      "step": 3405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0760862827301025,
+      "learning_rate": 4.910229976209358e-05,
+      "loss": 1.5566,
+      "step": 3406
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8208723068237305,
+      "learning_rate": 4.910203542162306e-05,
+      "loss": 1.5157,
+      "step": 3407
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6611407995224,
+      "learning_rate": 4.9101771081152525e-05,
+      "loss": 1.2632,
+      "step": 3408
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.720175266265869,
+      "learning_rate": 4.9101506740682005e-05,
+      "loss": 1.5497,
+      "step": 3409
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5092995166778564,
+      "learning_rate": 4.910124240021147e-05,
+      "loss": 1.2613,
+      "step": 3410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4808735847473145,
+      "learning_rate": 4.9100978059740945e-05,
+      "loss": 1.3575,
+      "step": 3411
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1173927783966064,
+      "learning_rate": 4.9100713719270426e-05,
+      "loss": 1.5111,
+      "step": 3412
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.738552212715149,
+      "learning_rate": 4.910044937879989e-05,
+      "loss": 1.4914,
+      "step": 3413
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.070114850997925,
+      "learning_rate": 4.910018503832937e-05,
+      "loss": 1.3561,
+      "step": 3414
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3695497512817383,
+      "learning_rate": 4.909992069785884e-05,
+      "loss": 1.3267,
+      "step": 3415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5164052248001099,
+      "learning_rate": 4.909965635738832e-05,
+      "loss": 1.3004,
+      "step": 3416
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5917686223983765,
+      "learning_rate": 4.9099392016917793e-05,
+      "loss": 1.2538,
+      "step": 3417
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5226860046386719,
+      "learning_rate": 4.909912767644727e-05,
+      "loss": 1.3997,
+      "step": 3418
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9368178844451904,
+      "learning_rate": 4.909886333597674e-05,
+      "loss": 1.5429,
+      "step": 3419
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7019696235656738,
+      "learning_rate": 4.9098598995506214e-05,
+      "loss": 1.3593,
+      "step": 3420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6061100959777832,
+      "learning_rate": 4.909833465503569e-05,
+      "loss": 1.473,
+      "step": 3421
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8612213134765625,
+      "learning_rate": 4.909807031456516e-05,
+      "loss": 1.3457,
+      "step": 3422
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.863876461982727,
+      "learning_rate": 4.9097805974094635e-05,
+      "loss": 1.4205,
+      "step": 3423
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7542575597763062,
+      "learning_rate": 4.909754163362411e-05,
+      "loss": 1.4636,
+      "step": 3424
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4493314027786255,
+      "learning_rate": 4.909727729315358e-05,
+      "loss": 1.3247,
+      "step": 3425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7078509330749512,
+      "learning_rate": 4.9097012952683056e-05,
+      "loss": 1.5183,
+      "step": 3426
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8109108209609985,
+      "learning_rate": 4.9096748612212536e-05,
+      "loss": 1.2838,
+      "step": 3427
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9151451587677002,
+      "learning_rate": 4.9096484271742e-05,
+      "loss": 1.3726,
+      "step": 3428
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.341580629348755,
+      "learning_rate": 4.909621993127148e-05,
+      "loss": 1.6529,
+      "step": 3429
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.461560606956482,
+      "learning_rate": 4.909595559080095e-05,
+      "loss": 1.3121,
+      "step": 3430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6430609226226807,
+      "learning_rate": 4.909569125033043e-05,
+      "loss": 1.3426,
+      "step": 3431
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1830496788024902,
+      "learning_rate": 4.9095426909859904e-05,
+      "loss": 1.335,
+      "step": 3432
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6726102828979492,
+      "learning_rate": 4.909516256938938e-05,
+      "loss": 1.402,
+      "step": 3433
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3736146688461304,
+      "learning_rate": 4.909489822891885e-05,
+      "loss": 1.3827,
+      "step": 3434
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7096070051193237,
+      "learning_rate": 4.9094633888448325e-05,
+      "loss": 1.5171,
+      "step": 3435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9434428215026855,
+      "learning_rate": 4.90943695479778e-05,
+      "loss": 1.479,
+      "step": 3436
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5344300270080566,
+      "learning_rate": 4.909410520750727e-05,
+      "loss": 1.5165,
+      "step": 3437
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.569554328918457,
+      "learning_rate": 4.9093840867036745e-05,
+      "loss": 1.4488,
+      "step": 3438
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6759605407714844,
+      "learning_rate": 4.909357652656622e-05,
+      "loss": 1.3106,
+      "step": 3439
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5370738506317139,
+      "learning_rate": 4.909331218609569e-05,
+      "loss": 1.4268,
+      "step": 3440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7749391794204712,
+      "learning_rate": 4.9093047845625166e-05,
+      "loss": 1.5672,
+      "step": 3441
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7914596796035767,
+      "learning_rate": 4.9092783505154646e-05,
+      "loss": 1.4963,
+      "step": 3442
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7004530429840088,
+      "learning_rate": 4.909251916468411e-05,
+      "loss": 1.5498,
+      "step": 3443
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7523000240325928,
+      "learning_rate": 4.9092254824213593e-05,
+      "loss": 1.5426,
+      "step": 3444
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.730065941810608,
+      "learning_rate": 4.909199048374306e-05,
+      "loss": 1.5167,
+      "step": 3445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6693766117095947,
+      "learning_rate": 4.909172614327254e-05,
+      "loss": 1.3852,
+      "step": 3446
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5312212705612183,
+      "learning_rate": 4.9091461802802014e-05,
+      "loss": 1.3831,
+      "step": 3447
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0785024166107178,
+      "learning_rate": 4.909119746233148e-05,
+      "loss": 1.4912,
+      "step": 3448
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8609364032745361,
+      "learning_rate": 4.909093312186096e-05,
+      "loss": 1.4708,
+      "step": 3449
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9385156631469727,
+      "learning_rate": 4.909066878139043e-05,
+      "loss": 1.5953,
+      "step": 3450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6912635564804077,
+      "learning_rate": 4.909040444091991e-05,
+      "loss": 1.372,
+      "step": 3451
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7073782682418823,
+      "learning_rate": 4.909014010044938e-05,
+      "loss": 1.4265,
+      "step": 3452
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.801315188407898,
+      "learning_rate": 4.9089875759978856e-05,
+      "loss": 1.466,
+      "step": 3453
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.781359314918518,
+      "learning_rate": 4.908961141950833e-05,
+      "loss": 1.3757,
+      "step": 3454
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4486321210861206,
+      "learning_rate": 4.90893470790378e-05,
+      "loss": 1.1441,
+      "step": 3455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3312455415725708,
+      "learning_rate": 4.9089082738567276e-05,
+      "loss": 1.2645,
+      "step": 3456
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4682685136795044,
+      "learning_rate": 4.908881839809675e-05,
+      "loss": 1.8298,
+      "step": 3457
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2888647317886353,
+      "learning_rate": 4.9088554057626223e-05,
+      "loss": 1.1369,
+      "step": 3458
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3667796850204468,
+      "learning_rate": 4.90882897171557e-05,
+      "loss": 1.3228,
+      "step": 3459
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7406567335128784,
+      "learning_rate": 4.908802537668517e-05,
+      "loss": 1.6361,
+      "step": 3460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7203922271728516,
+      "learning_rate": 4.9087761036214644e-05,
+      "loss": 1.4624,
+      "step": 3461
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.478483200073242,
+      "learning_rate": 4.9087496695744124e-05,
+      "loss": 1.5054,
+      "step": 3462
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9097323417663574,
+      "learning_rate": 4.908723235527359e-05,
+      "loss": 1.5223,
+      "step": 3463
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.694509744644165,
+      "learning_rate": 4.908696801480307e-05,
+      "loss": 1.4287,
+      "step": 3464
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.760381698608398,
+      "learning_rate": 4.908670367433254e-05,
+      "loss": 1.6715,
+      "step": 3465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.709886908531189,
+      "learning_rate": 4.908643933386202e-05,
+      "loss": 1.3865,
+      "step": 3466
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.622982144355774,
+      "learning_rate": 4.908617499339149e-05,
+      "loss": 1.3994,
+      "step": 3467
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6553770303726196,
+      "learning_rate": 4.9085910652920966e-05,
+      "loss": 1.3827,
+      "step": 3468
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.151766777038574,
+      "learning_rate": 4.908564631245044e-05,
+      "loss": 1.5398,
+      "step": 3469
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.171947479248047,
+      "learning_rate": 4.908538197197991e-05,
+      "loss": 1.3672,
+      "step": 3470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.683727502822876,
+      "learning_rate": 4.9085117631509387e-05,
+      "loss": 1.6465,
+      "step": 3471
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6418299674987793,
+      "learning_rate": 4.908485329103886e-05,
+      "loss": 1.3536,
+      "step": 3472
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7896841764450073,
+      "learning_rate": 4.9084588950568334e-05,
+      "loss": 1.3659,
+      "step": 3473
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5118001699447632,
+      "learning_rate": 4.908432461009781e-05,
+      "loss": 1.4249,
+      "step": 3474
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1021151542663574,
+      "learning_rate": 4.908406026962728e-05,
+      "loss": 2.3906,
+      "step": 3475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6728523969650269,
+      "learning_rate": 4.9083795929156754e-05,
+      "loss": 1.3948,
+      "step": 3476
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0980372428894043,
+      "learning_rate": 4.9083531588686235e-05,
+      "loss": 1.5533,
+      "step": 3477
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5856001377105713,
+      "learning_rate": 4.90832672482157e-05,
+      "loss": 1.4762,
+      "step": 3478
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4792900085449219,
+      "learning_rate": 4.908300290774518e-05,
+      "loss": 1.3752,
+      "step": 3479
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8008179664611816,
+      "learning_rate": 4.908273856727465e-05,
+      "loss": 1.9162,
+      "step": 3480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4514589309692383,
+      "learning_rate": 4.908247422680413e-05,
+      "loss": 1.4073,
+      "step": 3481
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7890101671218872,
+      "learning_rate": 4.90822098863336e-05,
+      "loss": 1.5113,
+      "step": 3482
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5512564182281494,
+      "learning_rate": 4.908194554586307e-05,
+      "loss": 1.6271,
+      "step": 3483
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.691638708114624,
+      "learning_rate": 4.908168120539255e-05,
+      "loss": 1.3703,
+      "step": 3484
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.299368381500244,
+      "learning_rate": 4.9081416864922017e-05,
+      "loss": 1.5445,
+      "step": 3485
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6993825435638428,
+      "learning_rate": 4.90811525244515e-05,
+      "loss": 1.4486,
+      "step": 3486
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6245325803756714,
+      "learning_rate": 4.908088818398097e-05,
+      "loss": 1.3324,
+      "step": 3487
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2477169036865234,
+      "learning_rate": 4.9080623843510444e-05,
+      "loss": 1.6472,
+      "step": 3488
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7021785974502563,
+      "learning_rate": 4.908035950303992e-05,
+      "loss": 1.4278,
+      "step": 3489
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6176952123641968,
+      "learning_rate": 4.908009516256939e-05,
+      "loss": 1.4302,
+      "step": 3490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4365471601486206,
+      "learning_rate": 4.9079830822098865e-05,
+      "loss": 1.407,
+      "step": 3491
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.604055166244507,
+      "learning_rate": 4.9079566481628345e-05,
+      "loss": 1.5225,
+      "step": 3492
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4055335521698,
+      "learning_rate": 4.907930214115781e-05,
+      "loss": 1.4972,
+      "step": 3493
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3643670082092285,
+      "learning_rate": 4.9079037800687286e-05,
+      "loss": 1.7405,
+      "step": 3494
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.687250018119812,
+      "learning_rate": 4.907877346021676e-05,
+      "loss": 1.3488,
+      "step": 3495
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7775986194610596,
+      "learning_rate": 4.907850911974623e-05,
+      "loss": 1.389,
+      "step": 3496
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.384979009628296,
+      "learning_rate": 4.907824477927571e-05,
+      "loss": 2.7159,
+      "step": 3497
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.416276693344116,
+      "learning_rate": 4.907798043880518e-05,
+      "loss": 1.5088,
+      "step": 3498
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6865066289901733,
+      "learning_rate": 4.907771609833466e-05,
+      "loss": 1.3325,
+      "step": 3499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4371752738952637,
+      "learning_rate": 4.907745175786413e-05,
+      "loss": 1.5631,
+      "step": 3500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7875920534133911,
+      "learning_rate": 4.907718741739361e-05,
+      "loss": 1.4436,
+      "step": 3501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9144178628921509,
+      "learning_rate": 4.907692307692308e-05,
+      "loss": 1.5412,
+      "step": 3502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.547636866569519,
+      "learning_rate": 4.9076658736452554e-05,
+      "loss": 2.0154,
+      "step": 3503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.130291223526001,
+      "learning_rate": 4.907639439598203e-05,
+      "loss": 1.3632,
+      "step": 3504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.830784797668457,
+      "learning_rate": 4.90761300555115e-05,
+      "loss": 1.2147,
+      "step": 3505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.479919910430908,
+      "learning_rate": 4.9075865715040975e-05,
+      "loss": 1.5294,
+      "step": 3506
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.563713788986206,
+      "learning_rate": 4.907560137457045e-05,
+      "loss": 1.442,
+      "step": 3507
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.962761402130127,
+      "learning_rate": 4.907533703409992e-05,
+      "loss": 1.4801,
+      "step": 3508
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.423513174057007,
+      "learning_rate": 4.9075072693629396e-05,
+      "loss": 1.5262,
+      "step": 3509
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6917424201965332,
+      "learning_rate": 4.907480835315887e-05,
+      "loss": 1.3704,
+      "step": 3510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8982964754104614,
+      "learning_rate": 4.907454401268834e-05,
+      "loss": 1.322,
+      "step": 3511
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0943362712860107,
+      "learning_rate": 4.907427967221782e-05,
+      "loss": 1.448,
+      "step": 3512
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7099437713623047,
+      "learning_rate": 4.907401533174729e-05,
+      "loss": 1.2215,
+      "step": 3513
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.601386070251465,
+      "learning_rate": 4.907375099127677e-05,
+      "loss": 1.3829,
+      "step": 3514
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4868420362472534,
+      "learning_rate": 4.907348665080624e-05,
+      "loss": 1.5635,
+      "step": 3515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7631601095199585,
+      "learning_rate": 4.907322231033572e-05,
+      "loss": 1.3026,
+      "step": 3516
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6458593606948853,
+      "learning_rate": 4.907295796986519e-05,
+      "loss": 1.4273,
+      "step": 3517
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7492479085922241,
+      "learning_rate": 4.9072693629394665e-05,
+      "loss": 1.5206,
+      "step": 3518
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.602602481842041,
+      "learning_rate": 4.907242928892414e-05,
+      "loss": 1.4115,
+      "step": 3519
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3813230991363525,
+      "learning_rate": 4.9072164948453605e-05,
+      "loss": 1.4272,
+      "step": 3520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5054874420166016,
+      "learning_rate": 4.9071900607983085e-05,
+      "loss": 1.3037,
+      "step": 3521
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.161731004714966,
+      "learning_rate": 4.907163626751256e-05,
+      "loss": 1.4995,
+      "step": 3522
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6367493867874146,
+      "learning_rate": 4.907137192704203e-05,
+      "loss": 1.5275,
+      "step": 3523
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.801389217376709,
+      "learning_rate": 4.9071107586571506e-05,
+      "loss": 1.468,
+      "step": 3524
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6683945655822754,
+      "learning_rate": 4.907084324610098e-05,
+      "loss": 1.3072,
+      "step": 3525
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5300241708755493,
+      "learning_rate": 4.907057890563045e-05,
+      "loss": 1.281,
+      "step": 3526
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.551286220550537,
+      "learning_rate": 4.9070314565159934e-05,
+      "loss": 1.3087,
+      "step": 3527
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.729437232017517,
+      "learning_rate": 4.90700502246894e-05,
+      "loss": 1.5469,
+      "step": 3528
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5769115686416626,
+      "learning_rate": 4.9069785884218874e-05,
+      "loss": 1.4937,
+      "step": 3529
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9697991609573364,
+      "learning_rate": 4.906952154374835e-05,
+      "loss": 1.4237,
+      "step": 3530
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9748010635375977,
+      "learning_rate": 4.906925720327782e-05,
+      "loss": 1.6428,
+      "step": 3531
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9651905298233032,
+      "learning_rate": 4.90689928628073e-05,
+      "loss": 1.3966,
+      "step": 3532
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0444881916046143,
+      "learning_rate": 4.906872852233677e-05,
+      "loss": 1.339,
+      "step": 3533
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6984232664108276,
+      "learning_rate": 4.906846418186625e-05,
+      "loss": 1.2974,
+      "step": 3534
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8484842777252197,
+      "learning_rate": 4.9068199841395715e-05,
+      "loss": 1.575,
+      "step": 3535
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3026610612869263,
+      "learning_rate": 4.9067935500925196e-05,
+      "loss": 1.3641,
+      "step": 3536
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0742123126983643,
+      "learning_rate": 4.906767116045467e-05,
+      "loss": 1.431,
+      "step": 3537
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9786494970321655,
+      "learning_rate": 4.906740681998414e-05,
+      "loss": 1.2611,
+      "step": 3538
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3670686483383179,
+      "learning_rate": 4.9067142479513617e-05,
+      "loss": 1.2599,
+      "step": 3539
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.514462947845459,
+      "learning_rate": 4.906687813904309e-05,
+      "loss": 1.3098,
+      "step": 3540
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.177983522415161,
+      "learning_rate": 4.9066613798572564e-05,
+      "loss": 1.5029,
+      "step": 3541
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7642035484313965,
+      "learning_rate": 4.906634945810204e-05,
+      "loss": 1.4882,
+      "step": 3542
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.268342018127441,
+      "learning_rate": 4.906608511763151e-05,
+      "loss": 1.569,
+      "step": 3543
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.286476969718933,
+      "learning_rate": 4.9065820777160984e-05,
+      "loss": 1.7921,
+      "step": 3544
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7122981548309326,
+      "learning_rate": 4.906555643669046e-05,
+      "loss": 1.5447,
+      "step": 3545
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5381722450256348,
+      "learning_rate": 4.906529209621993e-05,
+      "loss": 1.5208,
+      "step": 3546
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8347972631454468,
+      "learning_rate": 4.906502775574941e-05,
+      "loss": 1.5615,
+      "step": 3547
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8275872468948364,
+      "learning_rate": 4.906476341527888e-05,
+      "loss": 1.4014,
+      "step": 3548
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3988279104232788,
+      "learning_rate": 4.906449907480836e-05,
+      "loss": 1.3771,
+      "step": 3549
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0968244075775146,
+      "learning_rate": 4.9064234734337826e-05,
+      "loss": 1.3674,
+      "step": 3550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9244588613510132,
+      "learning_rate": 4.9063970393867306e-05,
+      "loss": 1.4461,
+      "step": 3551
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5717504024505615,
+      "learning_rate": 4.906370605339678e-05,
+      "loss": 1.4808,
+      "step": 3552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.732376217842102,
+      "learning_rate": 4.906344171292625e-05,
+      "loss": 1.5143,
+      "step": 3553
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8686282634735107,
+      "learning_rate": 4.906317737245573e-05,
+      "loss": 1.3114,
+      "step": 3554
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4218037128448486,
+      "learning_rate": 4.90629130319852e-05,
+      "loss": 1.656,
+      "step": 3555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.235902786254883,
+      "learning_rate": 4.9062648691514674e-05,
+      "loss": 1.2791,
+      "step": 3556
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2877671718597412,
+      "learning_rate": 4.906238435104415e-05,
+      "loss": 1.2026,
+      "step": 3557
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3037892580032349,
+      "learning_rate": 4.906212001057362e-05,
+      "loss": 2.2922,
+      "step": 3558
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1449625492095947,
+      "learning_rate": 4.9061855670103095e-05,
+      "loss": 1.4243,
+      "step": 3559
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6055399179458618,
+      "learning_rate": 4.906159132963257e-05,
+      "loss": 1.2507,
+      "step": 3560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8372706174850464,
+      "learning_rate": 4.906132698916204e-05,
+      "loss": 1.6292,
+      "step": 3561
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.103977918624878,
+      "learning_rate": 4.906106264869152e-05,
+      "loss": 1.373,
+      "step": 3562
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7152836322784424,
+      "learning_rate": 4.906079830822099e-05,
+      "loss": 1.49,
+      "step": 3563
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9187395572662354,
+      "learning_rate": 4.906053396775047e-05,
+      "loss": 1.4797,
+      "step": 3564
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.162980318069458,
+      "learning_rate": 4.9060269627279936e-05,
+      "loss": 1.3225,
+      "step": 3565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6922237873077393,
+      "learning_rate": 4.906000528680941e-05,
+      "loss": 1.3918,
+      "step": 3566
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.020923614501953,
+      "learning_rate": 4.905974094633889e-05,
+      "loss": 1.4673,
+      "step": 3567
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0734057426452637,
+      "learning_rate": 4.905947660586836e-05,
+      "loss": 1.609,
+      "step": 3568
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4142565727233887,
+      "learning_rate": 4.905921226539784e-05,
+      "loss": 1.3393,
+      "step": 3569
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7909032106399536,
+      "learning_rate": 4.9058947924927304e-05,
+      "loss": 1.2564,
+      "step": 3570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7372651100158691,
+      "learning_rate": 4.9058683584456784e-05,
+      "loss": 1.3041,
+      "step": 3571
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8122525215148926,
+      "learning_rate": 4.905841924398626e-05,
+      "loss": 1.8384,
+      "step": 3572
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8508553504943848,
+      "learning_rate": 4.905815490351573e-05,
+      "loss": 1.3357,
+      "step": 3573
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7850828170776367,
+      "learning_rate": 4.9057890563045205e-05,
+      "loss": 1.45,
+      "step": 3574
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.355427265167236,
+      "learning_rate": 4.905762622257468e-05,
+      "loss": 1.6909,
+      "step": 3575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8327348232269287,
+      "learning_rate": 4.905736188210415e-05,
+      "loss": 1.5518,
+      "step": 3576
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.793169379234314,
+      "learning_rate": 4.9057097541633626e-05,
+      "loss": 1.3152,
+      "step": 3577
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8846955299377441,
+      "learning_rate": 4.90568332011631e-05,
+      "loss": 1.2366,
+      "step": 3578
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3317697048187256,
+      "learning_rate": 4.905656886069257e-05,
+      "loss": 1.3512,
+      "step": 3579
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8246026039123535,
+      "learning_rate": 4.9056304520222046e-05,
+      "loss": 1.4414,
+      "step": 3580
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7181775569915771,
+      "learning_rate": 4.905604017975152e-05,
+      "loss": 1.153,
+      "step": 3581
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.215012311935425,
+      "learning_rate": 4.9055775839281e-05,
+      "loss": 1.3368,
+      "step": 3582
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5456879138946533,
+      "learning_rate": 4.905551149881047e-05,
+      "loss": 1.3674,
+      "step": 3583
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.705333709716797,
+      "learning_rate": 4.905524715833995e-05,
+      "loss": 1.6467,
+      "step": 3584
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2751379013061523,
+      "learning_rate": 4.9054982817869414e-05,
+      "loss": 1.5033,
+      "step": 3585
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9312559366226196,
+      "learning_rate": 4.9054718477398895e-05,
+      "loss": 1.5104,
+      "step": 3586
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9554845094680786,
+      "learning_rate": 4.905445413692837e-05,
+      "loss": 1.4005,
+      "step": 3587
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7533583641052246,
+      "learning_rate": 4.905418979645784e-05,
+      "loss": 1.4485,
+      "step": 3588
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.425510287284851,
+      "learning_rate": 4.9053925455987315e-05,
+      "loss": 1.644,
+      "step": 3589
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.493008017539978,
+      "learning_rate": 4.905366111551679e-05,
+      "loss": 1.3643,
+      "step": 3590
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4083389043807983,
+      "learning_rate": 4.905339677504626e-05,
+      "loss": 1.3283,
+      "step": 3591
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9644533395767212,
+      "learning_rate": 4.9053132434575736e-05,
+      "loss": 1.9062,
+      "step": 3592
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9601655006408691,
+      "learning_rate": 4.905286809410521e-05,
+      "loss": 2.0117,
+      "step": 3593
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.934584617614746,
+      "learning_rate": 4.905260375363468e-05,
+      "loss": 1.4401,
+      "step": 3594
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6890159845352173,
+      "learning_rate": 4.905233941316416e-05,
+      "loss": 1.4558,
+      "step": 3595
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4120630025863647,
+      "learning_rate": 4.905207507269363e-05,
+      "loss": 1.5851,
+      "step": 3596
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8829891681671143,
+      "learning_rate": 4.905181073222311e-05,
+      "loss": 1.5643,
+      "step": 3597
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.243126153945923,
+      "learning_rate": 4.905154639175258e-05,
+      "loss": 1.3944,
+      "step": 3598
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6958770751953125,
+      "learning_rate": 4.905128205128206e-05,
+      "loss": 1.3968,
+      "step": 3599
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.018549680709839,
+      "learning_rate": 4.9051017710811525e-05,
+      "loss": 1.3989,
+      "step": 3600
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0456175804138184,
+      "learning_rate": 4.9050753370341e-05,
+      "loss": 1.3827,
+      "step": 3601
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.226480722427368,
+      "learning_rate": 4.905048902987048e-05,
+      "loss": 1.4575,
+      "step": 3602
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7905023097991943,
+      "learning_rate": 4.9050224689399945e-05,
+      "loss": 1.4717,
+      "step": 3603
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9474985599517822,
+      "learning_rate": 4.9049960348929426e-05,
+      "loss": 1.4731,
+      "step": 3604
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6092896461486816,
+      "learning_rate": 4.904969600845889e-05,
+      "loss": 1.3805,
+      "step": 3605
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7488993406295776,
+      "learning_rate": 4.904943166798837e-05,
+      "loss": 1.3994,
+      "step": 3606
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.23021399974823,
+      "learning_rate": 4.9049167327517846e-05,
+      "loss": 1.137,
+      "step": 3607
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.562350273132324,
+      "learning_rate": 4.904890298704732e-05,
+      "loss": 1.5891,
+      "step": 3608
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5716358423233032,
+      "learning_rate": 4.9048638646576794e-05,
+      "loss": 1.3784,
+      "step": 3609
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.690797209739685,
+      "learning_rate": 4.904837430610627e-05,
+      "loss": 1.4274,
+      "step": 3610
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.238675117492676,
+      "learning_rate": 4.904810996563574e-05,
+      "loss": 1.4824,
+      "step": 3611
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5400443077087402,
+      "learning_rate": 4.9047845625165214e-05,
+      "loss": 1.3044,
+      "step": 3612
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5753371715545654,
+      "learning_rate": 4.904758128469469e-05,
+      "loss": 1.558,
+      "step": 3613
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4357213973999023,
+      "learning_rate": 4.904731694422416e-05,
+      "loss": 1.481,
+      "step": 3614
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.573021411895752,
+      "learning_rate": 4.9047052603753635e-05,
+      "loss": 1.3787,
+      "step": 3615
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6373435258865356,
+      "learning_rate": 4.904678826328311e-05,
+      "loss": 1.4854,
+      "step": 3616
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.969466209411621,
+      "learning_rate": 4.904652392281259e-05,
+      "loss": 1.3963,
+      "step": 3617
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.821424722671509,
+      "learning_rate": 4.9046259582342056e-05,
+      "loss": 1.8049,
+      "step": 3618
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.746992826461792,
+      "learning_rate": 4.9045995241871536e-05,
+      "loss": 1.2434,
+      "step": 3619
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.908360242843628,
+      "learning_rate": 4.9045730901401e-05,
+      "loss": 1.2945,
+      "step": 3620
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.772907018661499,
+      "learning_rate": 4.904546656093048e-05,
+      "loss": 2.3634,
+      "step": 3621
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.693406343460083,
+      "learning_rate": 4.904520222045996e-05,
+      "loss": 1.5061,
+      "step": 3622
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.181859254837036,
+      "learning_rate": 4.904493787998943e-05,
+      "loss": 1.6032,
+      "step": 3623
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1096031665802,
+      "learning_rate": 4.9044673539518904e-05,
+      "loss": 1.5273,
+      "step": 3624
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.850520372390747,
+      "learning_rate": 4.904440919904838e-05,
+      "loss": 1.5128,
+      "step": 3625
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.473235845565796,
+      "learning_rate": 4.904414485857785e-05,
+      "loss": 1.555,
+      "step": 3626
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.74553108215332,
+      "learning_rate": 4.9043880518107325e-05,
+      "loss": 2.1562,
+      "step": 3627
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8226159811019897,
+      "learning_rate": 4.90436161776368e-05,
+      "loss": 1.423,
+      "step": 3628
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6692842245101929,
+      "learning_rate": 4.904335183716627e-05,
+      "loss": 1.4121,
+      "step": 3629
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2997041940689087,
+      "learning_rate": 4.9043087496695745e-05,
+      "loss": 2.0885,
+      "step": 3630
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7339465618133545,
+      "learning_rate": 4.904282315622522e-05,
+      "loss": 1.2949,
+      "step": 3631
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7716742753982544,
+      "learning_rate": 4.90425588157547e-05,
+      "loss": 1.4934,
+      "step": 3632
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5220803022384644,
+      "learning_rate": 4.9042294475284166e-05,
+      "loss": 1.3034,
+      "step": 3633
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.848253846168518,
+      "learning_rate": 4.9042030134813646e-05,
+      "loss": 1.4647,
+      "step": 3634
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.774302363395691,
+      "learning_rate": 4.904176579434311e-05,
+      "loss": 1.3676,
+      "step": 3635
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5182603597640991,
+      "learning_rate": 4.9041501453872594e-05,
+      "loss": 1.32,
+      "step": 3636
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7409136295318604,
+      "learning_rate": 4.904123711340207e-05,
+      "loss": 1.5121,
+      "step": 3637
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3219513893127441,
+      "learning_rate": 4.9040972772931534e-05,
+      "loss": 1.2477,
+      "step": 3638
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6811139583587646,
+      "learning_rate": 4.9040708432461014e-05,
+      "loss": 1.3611,
+      "step": 3639
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1299917697906494,
+      "learning_rate": 4.904044409199048e-05,
+      "loss": 1.5857,
+      "step": 3640
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0441458225250244,
+      "learning_rate": 4.904017975151996e-05,
+      "loss": 1.5002,
+      "step": 3641
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4497742652893066,
+      "learning_rate": 4.9039915411049435e-05,
+      "loss": 1.3398,
+      "step": 3642
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5727852582931519,
+      "learning_rate": 4.903965107057891e-05,
+      "loss": 1.7301,
+      "step": 3643
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9773237705230713,
+      "learning_rate": 4.903938673010838e-05,
+      "loss": 1.5127,
+      "step": 3644
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.819098949432373,
+      "learning_rate": 4.9039122389637856e-05,
+      "loss": 1.3726,
+      "step": 3645
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8379641771316528,
+      "learning_rate": 4.903885804916733e-05,
+      "loss": 1.4916,
+      "step": 3646
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7515488862991333,
+      "learning_rate": 4.90385937086968e-05,
+      "loss": 1.4255,
+      "step": 3647
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1896555423736572,
+      "learning_rate": 4.9038329368226276e-05,
+      "loss": 1.4804,
+      "step": 3648
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1371066570281982,
+      "learning_rate": 4.903806502775575e-05,
+      "loss": 1.5125,
+      "step": 3649
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2750132083892822,
+      "learning_rate": 4.9037800687285224e-05,
+      "loss": 1.5441,
+      "step": 3650
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5235469341278076,
+      "learning_rate": 4.90375363468147e-05,
+      "loss": 1.5479,
+      "step": 3651
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4628403186798096,
+      "learning_rate": 4.903727200634418e-05,
+      "loss": 1.3956,
+      "step": 3652
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7767964601516724,
+      "learning_rate": 4.9037007665873644e-05,
+      "loss": 1.5708,
+      "step": 3653
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7068195343017578,
+      "learning_rate": 4.9036743325403125e-05,
+      "loss": 1.4446,
+      "step": 3654
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7720143795013428,
+      "learning_rate": 4.903647898493259e-05,
+      "loss": 1.5309,
+      "step": 3655
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4110312461853027,
+      "learning_rate": 4.903621464446207e-05,
+      "loss": 1.4832,
+      "step": 3656
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8272554874420166,
+      "learning_rate": 4.9035950303991545e-05,
+      "loss": 1.4719,
+      "step": 3657
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5961307287216187,
+      "learning_rate": 4.903568596352102e-05,
+      "loss": 1.3716,
+      "step": 3658
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.420744776725769,
+      "learning_rate": 4.903542162305049e-05,
+      "loss": 1.3291,
+      "step": 3659
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.439300775527954,
+      "learning_rate": 4.9035157282579966e-05,
+      "loss": 1.488,
+      "step": 3660
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.531050205230713,
+      "learning_rate": 4.903489294210944e-05,
+      "loss": 1.3364,
+      "step": 3661
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1445224285125732,
+      "learning_rate": 4.903462860163891e-05,
+      "loss": 1.4661,
+      "step": 3662
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4146511554718018,
+      "learning_rate": 4.903436426116839e-05,
+      "loss": 1.1931,
+      "step": 3663
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2828431129455566,
+      "learning_rate": 4.903409992069786e-05,
+      "loss": 1.3692,
+      "step": 3664
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8350026607513428,
+      "learning_rate": 4.9033835580227334e-05,
+      "loss": 1.3527,
+      "step": 3665
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3791559934616089,
+      "learning_rate": 4.903357123975681e-05,
+      "loss": 2.5904,
+      "step": 3666
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.459314227104187,
+      "learning_rate": 4.903330689928629e-05,
+      "loss": 1.3391,
+      "step": 3667
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9215986728668213,
+      "learning_rate": 4.9033042558815755e-05,
+      "loss": 1.4522,
+      "step": 3668
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.906765103340149,
+      "learning_rate": 4.9032778218345235e-05,
+      "loss": 1.5021,
+      "step": 3669
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2168482542037964,
+      "learning_rate": 4.90325138778747e-05,
+      "loss": 1.9734,
+      "step": 3670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6291898488998413,
+      "learning_rate": 4.903224953740418e-05,
+      "loss": 1.3802,
+      "step": 3671
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2550442218780518,
+      "learning_rate": 4.9031985196933656e-05,
+      "loss": 1.4043,
+      "step": 3672
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.496269702911377,
+      "learning_rate": 4.903172085646312e-05,
+      "loss": 1.4075,
+      "step": 3673
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7311818599700928,
+      "learning_rate": 4.90314565159926e-05,
+      "loss": 1.3742,
+      "step": 3674
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2772719860076904,
+      "learning_rate": 4.903119217552207e-05,
+      "loss": 1.2019,
+      "step": 3675
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5278011560440063,
+      "learning_rate": 4.903092783505155e-05,
+      "loss": 1.4739,
+      "step": 3676
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.745816946029663,
+      "learning_rate": 4.9030663494581024e-05,
+      "loss": 1.3674,
+      "step": 3677
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4598872661590576,
+      "learning_rate": 4.90303991541105e-05,
+      "loss": 1.3145,
+      "step": 3678
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.586664080619812,
+      "learning_rate": 4.903013481363997e-05,
+      "loss": 1.4706,
+      "step": 3679
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.668655514717102,
+      "learning_rate": 4.9029870473169444e-05,
+      "loss": 1.3921,
+      "step": 3680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6581472158432007,
+      "learning_rate": 4.902960613269892e-05,
+      "loss": 1.3987,
+      "step": 3681
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.759528398513794,
+      "learning_rate": 4.90293417922284e-05,
+      "loss": 1.3983,
+      "step": 3682
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3524270057678223,
+      "learning_rate": 4.9029077451757865e-05,
+      "loss": 1.27,
+      "step": 3683
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.762147307395935,
+      "learning_rate": 4.902881311128734e-05,
+      "loss": 1.6418,
+      "step": 3684
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.846395492553711,
+      "learning_rate": 4.902854877081681e-05,
+      "loss": 1.3617,
+      "step": 3685
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7751033306121826,
+      "learning_rate": 4.9028284430346286e-05,
+      "loss": 1.4368,
+      "step": 3686
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2203478813171387,
+      "learning_rate": 4.9028020089875766e-05,
+      "loss": 1.6313,
+      "step": 3687
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6050488948822021,
+      "learning_rate": 4.902775574940523e-05,
+      "loss": 1.299,
+      "step": 3688
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.285123109817505,
+      "learning_rate": 4.902749140893471e-05,
+      "loss": 1.4526,
+      "step": 3689
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1458773612976074,
+      "learning_rate": 4.902722706846418e-05,
+      "loss": 1.5076,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.766319990158081,
+      "learning_rate": 4.902696272799366e-05,
+      "loss": 1.4123,
+      "step": 3691
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8525747060775757,
+      "learning_rate": 4.9026698387523134e-05,
+      "loss": 1.3594,
+      "step": 3692
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.990355372428894,
+      "learning_rate": 4.902643404705261e-05,
+      "loss": 1.583,
+      "step": 3693
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8272980451583862,
+      "learning_rate": 4.902616970658208e-05,
+      "loss": 1.2405,
+      "step": 3694
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6033803224563599,
+      "learning_rate": 4.9025905366111555e-05,
+      "loss": 1.2887,
+      "step": 3695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0670251846313477,
+      "learning_rate": 4.902564102564103e-05,
+      "loss": 1.4034,
+      "step": 3696
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.909379482269287,
+      "learning_rate": 4.90253766851705e-05,
+      "loss": 1.3559,
+      "step": 3697
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5067005157470703,
+      "learning_rate": 4.9025112344699975e-05,
+      "loss": 1.5855,
+      "step": 3698
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9406238794326782,
+      "learning_rate": 4.902484800422945e-05,
+      "loss": 1.5357,
+      "step": 3699
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7827781438827515,
+      "learning_rate": 4.902458366375892e-05,
+      "loss": 1.362,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7741059064865112,
+      "learning_rate": 4.9024319323288396e-05,
+      "loss": 1.4788,
+      "step": 3701
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.419490098953247,
+      "learning_rate": 4.9024054982817876e-05,
+      "loss": 1.4079,
+      "step": 3702
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9614614248275757,
+      "learning_rate": 4.902379064234734e-05,
+      "loss": 1.4105,
+      "step": 3703
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8294812440872192,
+      "learning_rate": 4.9023526301876824e-05,
+      "loss": 2.2803,
+      "step": 3704
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.786432147026062,
+      "learning_rate": 4.902326196140629e-05,
+      "loss": 1.3816,
+      "step": 3705
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5392085313796997,
+      "learning_rate": 4.902299762093577e-05,
+      "loss": 1.3834,
+      "step": 3706
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3641347885131836,
+      "learning_rate": 4.9022733280465244e-05,
+      "loss": 1.4847,
+      "step": 3707
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.755181074142456,
+      "learning_rate": 4.902246893999472e-05,
+      "loss": 1.5923,
+      "step": 3708
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2361319065093994,
+      "learning_rate": 4.902220459952419e-05,
+      "loss": 1.4466,
+      "step": 3709
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8497642278671265,
+      "learning_rate": 4.902194025905366e-05,
+      "loss": 1.4964,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.457231879234314,
+      "learning_rate": 4.902167591858314e-05,
+      "loss": 1.4653,
+      "step": 3711
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.89536452293396,
+      "learning_rate": 4.902141157811261e-05,
+      "loss": 1.5017,
+      "step": 3712
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2121002674102783,
+      "learning_rate": 4.9021147237642086e-05,
+      "loss": 1.5913,
+      "step": 3713
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4894356727600098,
+      "learning_rate": 4.902088289717156e-05,
+      "loss": 1.5164,
+      "step": 3714
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5535852909088135,
+      "learning_rate": 4.902061855670103e-05,
+      "loss": 1.4452,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.649195671081543,
+      "learning_rate": 4.9020354216230506e-05,
+      "loss": 1.3641,
+      "step": 3716
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5573384761810303,
+      "learning_rate": 4.902008987575999e-05,
+      "loss": 1.3316,
+      "step": 3717
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8018426895141602,
+      "learning_rate": 4.9019825535289453e-05,
+      "loss": 1.4399,
+      "step": 3718
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8257180452346802,
+      "learning_rate": 4.901956119481893e-05,
+      "loss": 1.3697,
+      "step": 3719
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0061111450195312,
+      "learning_rate": 4.90192968543484e-05,
+      "loss": 1.5965,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6241426467895508,
+      "learning_rate": 4.9019032513877874e-05,
+      "loss": 1.4726,
+      "step": 3721
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.475329041481018,
+      "learning_rate": 4.9018768173407355e-05,
+      "loss": 1.3721,
+      "step": 3722
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3899650573730469,
+      "learning_rate": 4.901850383293682e-05,
+      "loss": 1.3656,
+      "step": 3723
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6987780332565308,
+      "learning_rate": 4.90182394924663e-05,
+      "loss": 1.5276,
+      "step": 3724
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2546863555908203,
+      "learning_rate": 4.901797515199577e-05,
+      "loss": 1.3377,
+      "step": 3725
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7785958051681519,
+      "learning_rate": 4.901771081152525e-05,
+      "loss": 1.2731,
+      "step": 3726
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9052908420562744,
+      "learning_rate": 4.901744647105472e-05,
+      "loss": 1.3663,
+      "step": 3727
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9903823137283325,
+      "learning_rate": 4.9017182130584196e-05,
+      "loss": 1.3701,
+      "step": 3728
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6779258251190186,
+      "learning_rate": 4.901691779011367e-05,
+      "loss": 1.5801,
+      "step": 3729
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.627333402633667,
+      "learning_rate": 4.901665344964314e-05,
+      "loss": 1.4043,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.698573350906372,
+      "learning_rate": 4.901638910917262e-05,
+      "loss": 1.4784,
+      "step": 3731
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7046316862106323,
+      "learning_rate": 4.901612476870209e-05,
+      "loss": 1.3974,
+      "step": 3732
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6788681745529175,
+      "learning_rate": 4.9015860428231564e-05,
+      "loss": 1.2894,
+      "step": 3733
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.87261164188385,
+      "learning_rate": 4.901559608776104e-05,
+      "loss": 1.3913,
+      "step": 3734
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.495726466178894,
+      "learning_rate": 4.901533174729051e-05,
+      "loss": 1.3539,
+      "step": 3735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.163137197494507,
+      "learning_rate": 4.9015067406819985e-05,
+      "loss": 1.3597,
+      "step": 3736
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8881237506866455,
+      "learning_rate": 4.9014803066349465e-05,
+      "loss": 1.3771,
+      "step": 3737
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2302345037460327,
+      "learning_rate": 4.901453872587893e-05,
+      "loss": 1.8735,
+      "step": 3738
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.917658805847168,
+      "learning_rate": 4.901427438540841e-05,
+      "loss": 1.5478,
+      "step": 3739
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6325674057006836,
+      "learning_rate": 4.901401004493788e-05,
+      "loss": 1.4082,
+      "step": 3740
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1978440284729004,
+      "learning_rate": 4.901374570446736e-05,
+      "loss": 1.3992,
+      "step": 3741
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.235567331314087,
+      "learning_rate": 4.901348136399683e-05,
+      "loss": 1.6558,
+      "step": 3742
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8530538082122803,
+      "learning_rate": 4.9013217023526306e-05,
+      "loss": 1.4716,
+      "step": 3743
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0268075466156006,
+      "learning_rate": 4.901295268305578e-05,
+      "loss": 1.3787,
+      "step": 3744
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9326603412628174,
+      "learning_rate": 4.9012688342585253e-05,
+      "loss": 1.2803,
+      "step": 3745
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.915601968765259,
+      "learning_rate": 4.901242400211473e-05,
+      "loss": 1.4774,
+      "step": 3746
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8211179971694946,
+      "learning_rate": 4.90121596616442e-05,
+      "loss": 1.2403,
+      "step": 3747
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9361273050308228,
+      "learning_rate": 4.9011895321173674e-05,
+      "loss": 1.5423,
+      "step": 3748
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.029754877090454,
+      "learning_rate": 4.901163098070315e-05,
+      "loss": 1.3529,
+      "step": 3749
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.613864541053772,
+      "learning_rate": 4.901136664023262e-05,
+      "loss": 1.4963,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4907883405685425,
+      "learning_rate": 4.9011102299762095e-05,
+      "loss": 1.3796,
+      "step": 3751
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4583786725997925,
+      "learning_rate": 4.9010837959291575e-05,
+      "loss": 1.418,
+      "step": 3752
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7544254064559937,
+      "learning_rate": 4.901057361882104e-05,
+      "loss": 1.4736,
+      "step": 3753
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7763404846191406,
+      "learning_rate": 4.901030927835052e-05,
+      "loss": 1.3732,
+      "step": 3754
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7532331943511963,
+      "learning_rate": 4.901004493787999e-05,
+      "loss": 1.4461,
+      "step": 3755
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.293867588043213,
+      "learning_rate": 4.900978059740946e-05,
+      "loss": 1.6347,
+      "step": 3756
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8184490203857422,
+      "learning_rate": 4.900951625693894e-05,
+      "loss": 1.3652,
+      "step": 3757
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9931488037109375,
+      "learning_rate": 4.900925191646841e-05,
+      "loss": 1.2826,
+      "step": 3758
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8889793157577515,
+      "learning_rate": 4.900898757599789e-05,
+      "loss": 1.5337,
+      "step": 3759
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7567137479782104,
+      "learning_rate": 4.900872323552736e-05,
+      "loss": 1.5025,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9439600706100464,
+      "learning_rate": 4.900845889505684e-05,
+      "loss": 1.5732,
+      "step": 3761
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8935198783874512,
+      "learning_rate": 4.900819455458631e-05,
+      "loss": 1.3066,
+      "step": 3762
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3897935152053833,
+      "learning_rate": 4.9007930214115785e-05,
+      "loss": 1.487,
+      "step": 3763
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4384649991989136,
+      "learning_rate": 4.900766587364526e-05,
+      "loss": 1.339,
+      "step": 3764
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.262401580810547,
+      "learning_rate": 4.900740153317473e-05,
+      "loss": 1.396,
+      "step": 3765
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7105013132095337,
+      "learning_rate": 4.9007137192704205e-05,
+      "loss": 1.5029,
+      "step": 3766
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.105731725692749,
+      "learning_rate": 4.900687285223368e-05,
+      "loss": 1.4215,
+      "step": 3767
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7798997163772583,
+      "learning_rate": 4.900660851176315e-05,
+      "loss": 1.4099,
+      "step": 3768
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4465258121490479,
+      "learning_rate": 4.9006344171292626e-05,
+      "loss": 1.4256,
+      "step": 3769
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8832892179489136,
+      "learning_rate": 4.90060798308221e-05,
+      "loss": 1.5759,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4802641868591309,
+      "learning_rate": 4.900581549035157e-05,
+      "loss": 1.3985,
+      "step": 3771
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0194156169891357,
+      "learning_rate": 4.9005551149881053e-05,
+      "loss": 1.5204,
+      "step": 3772
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9187575578689575,
+      "learning_rate": 4.900528680941052e-05,
+      "loss": 1.5297,
+      "step": 3773
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5197807550430298,
+      "learning_rate": 4.900502246894e-05,
+      "loss": 1.3889,
+      "step": 3774
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.566225290298462,
+      "learning_rate": 4.900475812846947e-05,
+      "loss": 1.414,
+      "step": 3775
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7736384868621826,
+      "learning_rate": 4.900449378799895e-05,
+      "loss": 1.4806,
+      "step": 3776
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.602662205696106,
+      "learning_rate": 4.900422944752842e-05,
+      "loss": 1.2687,
+      "step": 3777
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2320141792297363,
+      "learning_rate": 4.9003965107057895e-05,
+      "loss": 1.5389,
+      "step": 3778
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4285749197006226,
+      "learning_rate": 4.900370076658737e-05,
+      "loss": 1.2875,
+      "step": 3779
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.433940052986145,
+      "learning_rate": 4.900343642611684e-05,
+      "loss": 2.0644,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.28659188747406,
+      "learning_rate": 4.9003172085646316e-05,
+      "loss": 1.2414,
+      "step": 3781
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.04626202583313,
+      "learning_rate": 4.900290774517579e-05,
+      "loss": 1.4012,
+      "step": 3782
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.319114327430725,
+      "learning_rate": 4.900264340470526e-05,
+      "loss": 1.9613,
+      "step": 3783
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.330527424812317,
+      "learning_rate": 4.9002379064234736e-05,
+      "loss": 1.2675,
+      "step": 3784
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9076311588287354,
+      "learning_rate": 4.900211472376421e-05,
+      "loss": 1.21,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.547595739364624,
+      "learning_rate": 4.9001850383293683e-05,
+      "loss": 1.4601,
+      "step": 3786
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.995220422744751,
+      "learning_rate": 4.9001586042823164e-05,
+      "loss": 1.3236,
+      "step": 3787
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6233913898468018,
+      "learning_rate": 4.900132170235263e-05,
+      "loss": 1.3398,
+      "step": 3788
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3739830255508423,
+      "learning_rate": 4.900105736188211e-05,
+      "loss": 1.2262,
+      "step": 3789
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9948748350143433,
+      "learning_rate": 4.900079302141158e-05,
+      "loss": 1.5537,
+      "step": 3790
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7523120641708374,
+      "learning_rate": 4.900052868094105e-05,
+      "loss": 1.5222,
+      "step": 3791
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.823865532875061,
+      "learning_rate": 4.900026434047053e-05,
+      "loss": 1.5221,
+      "step": 3792
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0342304706573486,
+      "learning_rate": 4.9e-05,
+      "loss": 1.5358,
+      "step": 3793
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6722412109375,
+      "learning_rate": 4.899973565952948e-05,
+      "loss": 1.4944,
+      "step": 3794
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7506719827651978,
+      "learning_rate": 4.8999471319058946e-05,
+      "loss": 1.397,
+      "step": 3795
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.584514617919922,
+      "learning_rate": 4.8999206978588426e-05,
+      "loss": 2.3373,
+      "step": 3796
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8291599750518799,
+      "learning_rate": 4.89989426381179e-05,
+      "loss": 1.4243,
+      "step": 3797
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4779967069625854,
+      "learning_rate": 4.899867829764737e-05,
+      "loss": 1.1281,
+      "step": 3798
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.044718027114868,
+      "learning_rate": 4.899841395717685e-05,
+      "loss": 1.6468,
+      "step": 3799
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6296861171722412,
+      "learning_rate": 4.899814961670632e-05,
+      "loss": 1.4324,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8005307912826538,
+      "learning_rate": 4.8997885276235794e-05,
+      "loss": 1.4753,
+      "step": 3801
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.963437795639038,
+      "learning_rate": 4.899762093576527e-05,
+      "loss": 1.3757,
+      "step": 3802
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6247258186340332,
+      "learning_rate": 4.899735659529474e-05,
+      "loss": 1.2593,
+      "step": 3803
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.692556619644165,
+      "learning_rate": 4.8997092254824214e-05,
+      "loss": 1.5323,
+      "step": 3804
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.756623387336731,
+      "learning_rate": 4.899682791435369e-05,
+      "loss": 1.3979,
+      "step": 3805
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.264573335647583,
+      "learning_rate": 4.899656357388316e-05,
+      "loss": 1.3048,
+      "step": 3806
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8703848123550415,
+      "learning_rate": 4.899629923341264e-05,
+      "loss": 1.4554,
+      "step": 3807
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2075114250183105,
+      "learning_rate": 4.899603489294211e-05,
+      "loss": 1.5731,
+      "step": 3808
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7742152214050293,
+      "learning_rate": 4.899577055247159e-05,
+      "loss": 1.4433,
+      "step": 3809
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6616590023040771,
+      "learning_rate": 4.8995506212001056e-05,
+      "loss": 1.3986,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6017606258392334,
+      "learning_rate": 4.8995241871530536e-05,
+      "loss": 1.4566,
+      "step": 3811
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6157729625701904,
+      "learning_rate": 4.899497753106001e-05,
+      "loss": 1.6493,
+      "step": 3812
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7935621738433838,
+      "learning_rate": 4.8994713190589483e-05,
+      "loss": 1.6048,
+      "step": 3813
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3218860626220703,
+      "learning_rate": 4.899444885011896e-05,
+      "loss": 1.8107,
+      "step": 3814
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.519303560256958,
+      "learning_rate": 4.899418450964843e-05,
+      "loss": 1.4232,
+      "step": 3815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.662453293800354,
+      "learning_rate": 4.8993920169177904e-05,
+      "loss": 1.3219,
+      "step": 3816
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1972010135650635,
+      "learning_rate": 4.899365582870738e-05,
+      "loss": 1.7799,
+      "step": 3817
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7976939678192139,
+      "learning_rate": 4.899339148823685e-05,
+      "loss": 1.5626,
+      "step": 3818
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3675932884216309,
+      "learning_rate": 4.8993127147766325e-05,
+      "loss": 1.2554,
+      "step": 3819
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6567474603652954,
+      "learning_rate": 4.89928628072958e-05,
+      "loss": 1.3253,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.820319414138794,
+      "learning_rate": 4.899259846682527e-05,
+      "loss": 1.5103,
+      "step": 3821
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9529001712799072,
+      "learning_rate": 4.899233412635475e-05,
+      "loss": 1.5163,
+      "step": 3822
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4709845781326294,
+      "learning_rate": 4.899206978588422e-05,
+      "loss": 1.2174,
+      "step": 3823
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3322536945343018,
+      "learning_rate": 4.89918054454137e-05,
+      "loss": 1.255,
+      "step": 3824
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1479296684265137,
+      "learning_rate": 4.8991541104943166e-05,
+      "loss": 1.3958,
+      "step": 3825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5602099895477295,
+      "learning_rate": 4.8991276764472647e-05,
+      "loss": 1.3944,
+      "step": 3826
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4906059503555298,
+      "learning_rate": 4.899101242400212e-05,
+      "loss": 1.3921,
+      "step": 3827
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7099812030792236,
+      "learning_rate": 4.899074808353159e-05,
+      "loss": 1.5156,
+      "step": 3828
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0248661041259766,
+      "learning_rate": 4.899048374306107e-05,
+      "loss": 1.5445,
+      "step": 3829
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.773000955581665,
+      "learning_rate": 4.8990219402590534e-05,
+      "loss": 1.3875,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4110591411590576,
+      "learning_rate": 4.8989955062120014e-05,
+      "loss": 2.0797,
+      "step": 3831
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6612275838851929,
+      "learning_rate": 4.898969072164949e-05,
+      "loss": 1.4122,
+      "step": 3832
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7331550121307373,
+      "learning_rate": 4.898942638117896e-05,
+      "loss": 1.4337,
+      "step": 3833
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7885103225708008,
+      "learning_rate": 4.8989162040708435e-05,
+      "loss": 1.4441,
+      "step": 3834
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6837012767791748,
+      "learning_rate": 4.898889770023791e-05,
+      "loss": 1.5078,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0805370807647705,
+      "learning_rate": 4.898863335976738e-05,
+      "loss": 2.3644,
+      "step": 3836
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1700143814086914,
+      "learning_rate": 4.8988369019296856e-05,
+      "loss": 1.535,
+      "step": 3837
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7645186185836792,
+      "learning_rate": 4.898810467882633e-05,
+      "loss": 1.9001,
+      "step": 3838
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6367549896240234,
+      "learning_rate": 4.89878403383558e-05,
+      "loss": 1.337,
+      "step": 3839
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5657918453216553,
+      "learning_rate": 4.8987575997885277e-05,
+      "loss": 1.3876,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.310001850128174,
+      "learning_rate": 4.898731165741475e-05,
+      "loss": 1.3637,
+      "step": 3841
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4661879539489746,
+      "learning_rate": 4.898704731694423e-05,
+      "loss": 1.4574,
+      "step": 3842
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0144200325012207,
+      "learning_rate": 4.89867829764737e-05,
+      "loss": 1.5387,
+      "step": 3843
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.532204508781433,
+      "learning_rate": 4.898651863600318e-05,
+      "loss": 1.3341,
+      "step": 3844
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0932743549346924,
+      "learning_rate": 4.8986254295532644e-05,
+      "loss": 1.467,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4919214248657227,
+      "learning_rate": 4.8985989955062125e-05,
+      "loss": 1.5815,
+      "step": 3846
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5880374908447266,
+      "learning_rate": 4.89857256145916e-05,
+      "loss": 1.3871,
+      "step": 3847
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.683620572090149,
+      "learning_rate": 4.898546127412107e-05,
+      "loss": 1.3897,
+      "step": 3848
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0589029788970947,
+      "learning_rate": 4.8985196933650545e-05,
+      "loss": 1.5098,
+      "step": 3849
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5419204235076904,
+      "learning_rate": 4.898493259318002e-05,
+      "loss": 1.2326,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4621365070343018,
+      "learning_rate": 4.898466825270949e-05,
+      "loss": 1.4287,
+      "step": 3851
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5100780725479126,
+      "learning_rate": 4.8984403912238966e-05,
+      "loss": 1.3937,
+      "step": 3852
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.273653984069824,
+      "learning_rate": 4.898413957176844e-05,
+      "loss": 1.6194,
+      "step": 3853
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7931722402572632,
+      "learning_rate": 4.898387523129791e-05,
+      "loss": 1.4927,
+      "step": 3854
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8550350666046143,
+      "learning_rate": 4.898361089082739e-05,
+      "loss": 1.5898,
+      "step": 3855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0095067024230957,
+      "learning_rate": 4.898334655035686e-05,
+      "loss": 1.5478,
+      "step": 3856
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8124957084655762,
+      "learning_rate": 4.898308220988634e-05,
+      "loss": 1.495,
+      "step": 3857
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.68497633934021,
+      "learning_rate": 4.898281786941581e-05,
+      "loss": 1.5196,
+      "step": 3858
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3836697340011597,
+      "learning_rate": 4.898255352894529e-05,
+      "loss": 1.618,
+      "step": 3859
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.47968590259552,
+      "learning_rate": 4.8982289188474755e-05,
+      "loss": 1.356,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1337363719940186,
+      "learning_rate": 4.8982024848004235e-05,
+      "loss": 1.5549,
+      "step": 3861
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5475281476974487,
+      "learning_rate": 4.898176050753371e-05,
+      "loss": 1.3836,
+      "step": 3862
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5405023097991943,
+      "learning_rate": 4.898149616706318e-05,
+      "loss": 1.3084,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7821701765060425,
+      "learning_rate": 4.8981231826592656e-05,
+      "loss": 1.4766,
+      "step": 3864
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7633503675460815,
+      "learning_rate": 4.898096748612212e-05,
+      "loss": 1.5409,
+      "step": 3865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5785460472106934,
+      "learning_rate": 4.89807031456516e-05,
+      "loss": 1.3711,
+      "step": 3866
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5835256576538086,
+      "learning_rate": 4.8980438805181077e-05,
+      "loss": 1.4199,
+      "step": 3867
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6997792720794678,
+      "learning_rate": 4.898017446471055e-05,
+      "loss": 1.3309,
+      "step": 3868
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.919049620628357,
+      "learning_rate": 4.8979910124240024e-05,
+      "loss": 1.518,
+      "step": 3869
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8451262712478638,
+      "learning_rate": 4.89796457837695e-05,
+      "loss": 1.52,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7188917398452759,
+      "learning_rate": 4.897938144329897e-05,
+      "loss": 1.2807,
+      "step": 3871
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7366305589675903,
+      "learning_rate": 4.897911710282845e-05,
+      "loss": 1.406,
+      "step": 3872
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7411003112792969,
+      "learning_rate": 4.897885276235792e-05,
+      "loss": 1.4015,
+      "step": 3873
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9482814073562622,
+      "learning_rate": 4.897858842188739e-05,
+      "loss": 1.5082,
+      "step": 3874
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1530351638793945,
+      "learning_rate": 4.8978324081416865e-05,
+      "loss": 1.4032,
+      "step": 3875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4238834381103516,
+      "learning_rate": 4.897805974094634e-05,
+      "loss": 1.4071,
+      "step": 3876
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.150369167327881,
+      "learning_rate": 4.897779540047582e-05,
+      "loss": 1.4427,
+      "step": 3877
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8796385526657104,
+      "learning_rate": 4.8977531060005286e-05,
+      "loss": 1.3298,
+      "step": 3878
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.534011721611023,
+      "learning_rate": 4.8977266719534766e-05,
+      "loss": 1.4375,
+      "step": 3879
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4692766666412354,
+      "learning_rate": 4.897700237906423e-05,
+      "loss": 1.3532,
+      "step": 3880
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5724612474441528,
+      "learning_rate": 4.897673803859371e-05,
+      "loss": 1.376,
+      "step": 3881
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.121776580810547,
+      "learning_rate": 4.897647369812319e-05,
+      "loss": 1.6172,
+      "step": 3882
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7559559345245361,
+      "learning_rate": 4.897620935765266e-05,
+      "loss": 1.3658,
+      "step": 3883
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6748439073562622,
+      "learning_rate": 4.8975945017182134e-05,
+      "loss": 1.343,
+      "step": 3884
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5395309925079346,
+      "learning_rate": 4.897568067671161e-05,
+      "loss": 1.3104,
+      "step": 3885
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7587451934814453,
+      "learning_rate": 4.897541633624108e-05,
+      "loss": 1.6188,
+      "step": 3886
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2649576663970947,
+      "learning_rate": 4.8975151995770555e-05,
+      "loss": 1.4447,
+      "step": 3887
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.000525951385498,
+      "learning_rate": 4.897488765530003e-05,
+      "loss": 1.3934,
+      "step": 3888
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6648173332214355,
+      "learning_rate": 4.89746233148295e-05,
+      "loss": 1.4996,
+      "step": 3889
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3375864028930664,
+      "learning_rate": 4.8974358974358975e-05,
+      "loss": 1.4754,
+      "step": 3890
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0975873470306396,
+      "learning_rate": 4.897409463388845e-05,
+      "loss": 1.4354,
+      "step": 3891
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8878519535064697,
+      "learning_rate": 4.897383029341793e-05,
+      "loss": 1.5408,
+      "step": 3892
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.109114408493042,
+      "learning_rate": 4.8973565952947396e-05,
+      "loss": 1.4135,
+      "step": 3893
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8337503671646118,
+      "learning_rate": 4.8973301612476877e-05,
+      "loss": 1.4244,
+      "step": 3894
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7370907068252563,
+      "learning_rate": 4.897303727200634e-05,
+      "loss": 1.4011,
+      "step": 3895
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6897159814834595,
+      "learning_rate": 4.8972772931535824e-05,
+      "loss": 2.3695,
+      "step": 3896
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7793078422546387,
+      "learning_rate": 4.89725085910653e-05,
+      "loss": 1.3473,
+      "step": 3897
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.843485713005066,
+      "learning_rate": 4.897224425059477e-05,
+      "loss": 1.3306,
+      "step": 3898
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.019906759262085,
+      "learning_rate": 4.8971979910124244e-05,
+      "loss": 1.5342,
+      "step": 3899
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1599106788635254,
+      "learning_rate": 4.897171556965371e-05,
+      "loss": 1.5121,
+      "step": 3900
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.918248176574707,
+      "learning_rate": 4.897145122918319e-05,
+      "loss": 1.4187,
+      "step": 3901
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8706752061843872,
+      "learning_rate": 4.8971186888712665e-05,
+      "loss": 1.4649,
+      "step": 3902
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9147380590438843,
+      "learning_rate": 4.897092254824214e-05,
+      "loss": 1.444,
+      "step": 3903
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3371425867080688,
+      "learning_rate": 4.897065820777161e-05,
+      "loss": 1.2851,
+      "step": 3904
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6749950647354126,
+      "learning_rate": 4.8970393867301086e-05,
+      "loss": 1.4037,
+      "step": 3905
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.855055332183838,
+      "learning_rate": 4.897012952683056e-05,
+      "loss": 1.4305,
+      "step": 3906
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4792581796646118,
+      "learning_rate": 4.896986518636004e-05,
+      "loss": 1.3489,
+      "step": 3907
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6368824243545532,
+      "learning_rate": 4.8969600845889507e-05,
+      "loss": 1.4347,
+      "step": 3908
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8454145193099976,
+      "learning_rate": 4.896933650541898e-05,
+      "loss": 1.3776,
+      "step": 3909
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5830799341201782,
+      "learning_rate": 4.8969072164948454e-05,
+      "loss": 2.4635,
+      "step": 3910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9399181604385376,
+      "learning_rate": 4.896880782447793e-05,
+      "loss": 1.5168,
+      "step": 3911
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9537070989608765,
+      "learning_rate": 4.896854348400741e-05,
+      "loss": 1.5313,
+      "step": 3912
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5051463842391968,
+      "learning_rate": 4.8968279143536874e-05,
+      "loss": 1.3471,
+      "step": 3913
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.912665605545044,
+      "learning_rate": 4.8968014803066355e-05,
+      "loss": 1.4175,
+      "step": 3914
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.172377347946167,
+      "learning_rate": 4.896775046259582e-05,
+      "loss": 1.5552,
+      "step": 3915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.642593502998352,
+      "learning_rate": 4.89674861221253e-05,
+      "loss": 1.3861,
+      "step": 3916
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4684878587722778,
+      "learning_rate": 4.8967221781654775e-05,
+      "loss": 1.3176,
+      "step": 3917
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.076301336288452,
+      "learning_rate": 4.896695744118425e-05,
+      "loss": 1.2498,
+      "step": 3918
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9841138124465942,
+      "learning_rate": 4.896669310071372e-05,
+      "loss": 1.5223,
+      "step": 3919
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8598560094833374,
+      "learning_rate": 4.8966428760243196e-05,
+      "loss": 1.5949,
+      "step": 3920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4974629878997803,
+      "learning_rate": 4.896616441977267e-05,
+      "loss": 1.6191,
+      "step": 3921
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5819873809814453,
+      "learning_rate": 4.896590007930214e-05,
+      "loss": 1.3663,
+      "step": 3922
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5862518548965454,
+      "learning_rate": 4.896563573883162e-05,
+      "loss": 1.3702,
+      "step": 3923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4954396486282349,
+      "learning_rate": 4.896537139836109e-05,
+      "loss": 1.3288,
+      "step": 3924
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.000331401824951,
+      "learning_rate": 4.8965107057890564e-05,
+      "loss": 1.4223,
+      "step": 3925
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6508498191833496,
+      "learning_rate": 4.896484271742004e-05,
+      "loss": 1.4465,
+      "step": 3926
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.680675983428955,
+      "learning_rate": 4.896457837694952e-05,
+      "loss": 1.5309,
+      "step": 3927
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4532872438430786,
+      "learning_rate": 4.8964314036478985e-05,
+      "loss": 1.338,
+      "step": 3928
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6584159135818481,
+      "learning_rate": 4.8964049696008465e-05,
+      "loss": 1.3868,
+      "step": 3929
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5322322845458984,
+      "learning_rate": 4.896378535553793e-05,
+      "loss": 1.3085,
+      "step": 3930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0201709270477295,
+      "learning_rate": 4.896352101506741e-05,
+      "loss": 1.6329,
+      "step": 3931
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7089500427246094,
+      "learning_rate": 4.8963256674596886e-05,
+      "loss": 1.2577,
+      "step": 3932
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.164731025695801,
+      "learning_rate": 4.896299233412636e-05,
+      "loss": 1.4397,
+      "step": 3933
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5506772994995117,
+      "learning_rate": 4.896272799365583e-05,
+      "loss": 1.4607,
+      "step": 3934
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6369359493255615,
+      "learning_rate": 4.8962463653185306e-05,
+      "loss": 1.4207,
+      "step": 3935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6959437131881714,
+      "learning_rate": 4.896219931271478e-05,
+      "loss": 1.4275,
+      "step": 3936
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.001929998397827,
+      "learning_rate": 4.8961934972244254e-05,
+      "loss": 1.5593,
+      "step": 3937
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2277939319610596,
+      "learning_rate": 4.896167063177373e-05,
+      "loss": 1.2678,
+      "step": 3938
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8015168905258179,
+      "learning_rate": 4.89614062913032e-05,
+      "loss": 1.4498,
+      "step": 3939
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.466732144355774,
+      "learning_rate": 4.8961141950832674e-05,
+      "loss": 1.2962,
+      "step": 3940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1445224285125732,
+      "learning_rate": 4.896087761036215e-05,
+      "loss": 1.4992,
+      "step": 3941
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5271073579788208,
+      "learning_rate": 4.896061326989163e-05,
+      "loss": 1.4167,
+      "step": 3942
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.919288158416748,
+      "learning_rate": 4.8960348929421095e-05,
+      "loss": 1.2447,
+      "step": 3943
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.175103187561035,
+      "learning_rate": 4.8960084588950575e-05,
+      "loss": 1.4955,
+      "step": 3944
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.565294861793518,
+      "learning_rate": 4.895982024848004e-05,
+      "loss": 1.4173,
+      "step": 3945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2089788913726807,
+      "learning_rate": 4.8959555908009516e-05,
+      "loss": 1.589,
+      "step": 3946
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7738690376281738,
+      "learning_rate": 4.8959291567538996e-05,
+      "loss": 1.3842,
+      "step": 3947
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6989868879318237,
+      "learning_rate": 4.895902722706846e-05,
+      "loss": 1.4092,
+      "step": 3948
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6856032609939575,
+      "learning_rate": 4.895876288659794e-05,
+      "loss": 1.3649,
+      "step": 3949
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.010573387145996,
+      "learning_rate": 4.895849854612741e-05,
+      "loss": 1.566,
+      "step": 3950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8344911336898804,
+      "learning_rate": 4.895823420565689e-05,
+      "loss": 1.4479,
+      "step": 3951
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9264808893203735,
+      "learning_rate": 4.8957969865186364e-05,
+      "loss": 1.401,
+      "step": 3952
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7237566709518433,
+      "learning_rate": 4.895770552471584e-05,
+      "loss": 1.406,
+      "step": 3953
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8245798349380493,
+      "learning_rate": 4.895744118424531e-05,
+      "loss": 1.3665,
+      "step": 3954
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.722141981124878,
+      "learning_rate": 4.8957176843774785e-05,
+      "loss": 1.248,
+      "step": 3955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.018702507019043,
+      "learning_rate": 4.895691250330426e-05,
+      "loss": 1.5532,
+      "step": 3956
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4392720460891724,
+      "learning_rate": 4.895664816283373e-05,
+      "loss": 1.3376,
+      "step": 3957
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.146930456161499,
+      "learning_rate": 4.8956383822363205e-05,
+      "loss": 1.3549,
+      "step": 3958
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.068838119506836,
+      "learning_rate": 4.895611948189268e-05,
+      "loss": 1.3822,
+      "step": 3959
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.779256820678711,
+      "learning_rate": 4.895585514142215e-05,
+      "loss": 1.457,
+      "step": 3960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7320919036865234,
+      "learning_rate": 4.8955590800951626e-05,
+      "loss": 1.5288,
+      "step": 3961
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7974258661270142,
+      "learning_rate": 4.8955326460481106e-05,
+      "loss": 1.5479,
+      "step": 3962
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5383894443511963,
+      "learning_rate": 4.895506212001057e-05,
+      "loss": 1.456,
+      "step": 3963
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9763518571853638,
+      "learning_rate": 4.8954797779540054e-05,
+      "loss": 1.2127,
+      "step": 3964
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2327663898468018,
+      "learning_rate": 4.895453343906952e-05,
+      "loss": 1.4413,
+      "step": 3965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3217577934265137,
+      "learning_rate": 4.8954269098599e-05,
+      "loss": 1.6641,
+      "step": 3966
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.386179208755493,
+      "learning_rate": 4.8954004758128474e-05,
+      "loss": 1.358,
+      "step": 3967
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4433059692382812,
+      "learning_rate": 4.895374041765795e-05,
+      "loss": 1.5001,
+      "step": 3968
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9387989044189453,
+      "learning_rate": 4.895347607718742e-05,
+      "loss": 1.4096,
+      "step": 3969
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7496109008789062,
+      "learning_rate": 4.8953211736716895e-05,
+      "loss": 1.5044,
+      "step": 3970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9423604011535645,
+      "learning_rate": 4.895294739624637e-05,
+      "loss": 1.3397,
+      "step": 3971
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9741805791854858,
+      "learning_rate": 4.895268305577584e-05,
+      "loss": 1.624,
+      "step": 3972
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5975502729415894,
+      "learning_rate": 4.8952418715305316e-05,
+      "loss": 1.4449,
+      "step": 3973
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6909186840057373,
+      "learning_rate": 4.895215437483479e-05,
+      "loss": 1.4357,
+      "step": 3974
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.613282322883606,
+      "learning_rate": 4.895189003436426e-05,
+      "loss": 1.2458,
+      "step": 3975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.980503559112549,
+      "learning_rate": 4.8951625693893736e-05,
+      "loss": 1.3671,
+      "step": 3976
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4917957782745361,
+      "learning_rate": 4.895136135342322e-05,
+      "loss": 1.4834,
+      "step": 3977
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9776612520217896,
+      "learning_rate": 4.8951097012952684e-05,
+      "loss": 1.7069,
+      "step": 3978
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6148180961608887,
+      "learning_rate": 4.8950832672482164e-05,
+      "loss": 1.5517,
+      "step": 3979
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.635842204093933,
+      "learning_rate": 4.895056833201163e-05,
+      "loss": 1.4203,
+      "step": 3980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5660372972488403,
+      "learning_rate": 4.895030399154111e-05,
+      "loss": 1.3538,
+      "step": 3981
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9096733331680298,
+      "learning_rate": 4.8950039651070585e-05,
+      "loss": 1.4059,
+      "step": 3982
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8798021078109741,
+      "learning_rate": 4.894977531060005e-05,
+      "loss": 1.4994,
+      "step": 3983
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.582588791847229,
+      "learning_rate": 4.894951097012953e-05,
+      "loss": 1.3203,
+      "step": 3984
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9254887104034424,
+      "learning_rate": 4.8949246629659e-05,
+      "loss": 1.457,
+      "step": 3985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.447129011154175,
+      "learning_rate": 4.894898228918848e-05,
+      "loss": 1.5089,
+      "step": 3986
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5117042064666748,
+      "learning_rate": 4.894871794871795e-05,
+      "loss": 1.5022,
+      "step": 3987
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.673671007156372,
+      "learning_rate": 4.8948453608247426e-05,
+      "loss": 1.3916,
+      "step": 3988
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.65035080909729,
+      "learning_rate": 4.89481892677769e-05,
+      "loss": 1.405,
+      "step": 3989
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5610108375549316,
+      "learning_rate": 4.894792492730637e-05,
+      "loss": 1.3115,
+      "step": 3990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9665498733520508,
+      "learning_rate": 4.894766058683585e-05,
+      "loss": 1.4446,
+      "step": 3991
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.49069344997406,
+      "learning_rate": 4.894739624636532e-05,
+      "loss": 1.3287,
+      "step": 3992
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.321232795715332,
+      "learning_rate": 4.8947131905894794e-05,
+      "loss": 1.2633,
+      "step": 3993
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4515419006347656,
+      "learning_rate": 4.894686756542427e-05,
+      "loss": 1.5351,
+      "step": 3994
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5308115482330322,
+      "learning_rate": 4.894660322495374e-05,
+      "loss": 1.5213,
+      "step": 3995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4023276567459106,
+      "learning_rate": 4.8946338884483215e-05,
+      "loss": 1.2981,
+      "step": 3996
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6036696434020996,
+      "learning_rate": 4.8946074544012695e-05,
+      "loss": 1.4001,
+      "step": 3997
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.911238431930542,
+      "learning_rate": 4.894581020354216e-05,
+      "loss": 1.4947,
+      "step": 3998
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7973638772964478,
+      "learning_rate": 4.894554586307164e-05,
+      "loss": 1.4699,
+      "step": 3999
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5897597074508667,
+      "learning_rate": 4.894528152260111e-05,
+      "loss": 1.4517,
+      "step": 4000
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6060668230056763,
+      "learning_rate": 4.894501718213059e-05,
+      "loss": 1.3486,
+      "step": 4001
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9673707485198975,
+      "learning_rate": 4.894475284166006e-05,
+      "loss": 1.3759,
+      "step": 4002
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6685044765472412,
+      "learning_rate": 4.8944488501189536e-05,
+      "loss": 1.4399,
+      "step": 4003
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8136906623840332,
+      "learning_rate": 4.894422416071901e-05,
+      "loss": 1.5206,
+      "step": 4004
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7765573263168335,
+      "learning_rate": 4.8943959820248484e-05,
+      "loss": 1.4376,
+      "step": 4005
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.268601417541504,
+      "learning_rate": 4.894369547977796e-05,
+      "loss": 1.5917,
+      "step": 4006
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.146359920501709,
+      "learning_rate": 4.894343113930743e-05,
+      "loss": 1.5445,
+      "step": 4007
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4715282917022705,
+      "learning_rate": 4.8943166798836904e-05,
+      "loss": 1.3875,
+      "step": 4008
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.648827075958252,
+      "learning_rate": 4.894290245836638e-05,
+      "loss": 1.3931,
+      "step": 4009
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5960155725479126,
+      "learning_rate": 4.894263811789585e-05,
+      "loss": 1.5418,
+      "step": 4010
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0991411209106445,
+      "learning_rate": 4.8942373777425325e-05,
+      "loss": 1.6401,
+      "step": 4011
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.964232087135315,
+      "learning_rate": 4.8942109436954805e-05,
+      "loss": 1.4066,
+      "step": 4012
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5524553060531616,
+      "learning_rate": 4.894184509648427e-05,
+      "loss": 1.2258,
+      "step": 4013
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8086827993392944,
+      "learning_rate": 4.894158075601375e-05,
+      "loss": 1.553,
+      "step": 4014
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5223890542984009,
+      "learning_rate": 4.894131641554322e-05,
+      "loss": 1.1772,
+      "step": 4015
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.093034029006958,
+      "learning_rate": 4.89410520750727e-05,
+      "loss": 1.4122,
+      "step": 4016
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5530517101287842,
+      "learning_rate": 4.894078773460217e-05,
+      "loss": 1.2925,
+      "step": 4017
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6001044511795044,
+      "learning_rate": 4.894052339413164e-05,
+      "loss": 1.4835,
+      "step": 4018
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.266371488571167,
+      "learning_rate": 4.894025905366112e-05,
+      "loss": 1.5902,
+      "step": 4019
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3782769441604614,
+      "learning_rate": 4.893999471319059e-05,
+      "loss": 1.1045,
+      "step": 4020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.750001311302185,
+      "learning_rate": 4.893973037272007e-05,
+      "loss": 1.5231,
+      "step": 4021
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9406734704971313,
+      "learning_rate": 4.893946603224954e-05,
+      "loss": 1.4026,
+      "step": 4022
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.619444489479065,
+      "learning_rate": 4.8939201691779015e-05,
+      "loss": 1.3826,
+      "step": 4023
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9563277959823608,
+      "learning_rate": 4.893893735130849e-05,
+      "loss": 1.5003,
+      "step": 4024
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7892887592315674,
+      "learning_rate": 4.893867301083796e-05,
+      "loss": 1.501,
+      "step": 4025
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9329617023468018,
+      "learning_rate": 4.8938408670367435e-05,
+      "loss": 1.5528,
+      "step": 4026
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6203283071517944,
+      "learning_rate": 4.893814432989691e-05,
+      "loss": 1.3972,
+      "step": 4027
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9410659074783325,
+      "learning_rate": 4.893787998942638e-05,
+      "loss": 1.4649,
+      "step": 4028
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4861202239990234,
+      "learning_rate": 4.8937615648955856e-05,
+      "loss": 1.5771,
+      "step": 4029
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.19840407371521,
+      "learning_rate": 4.893735130848533e-05,
+      "loss": 1.3815,
+      "step": 4030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7819507122039795,
+      "learning_rate": 4.89370869680148e-05,
+      "loss": 1.3911,
+      "step": 4031
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5804696083068848,
+      "learning_rate": 4.8936822627544284e-05,
+      "loss": 1.3263,
+      "step": 4032
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5018980503082275,
+      "learning_rate": 4.893655828707375e-05,
+      "loss": 1.4347,
+      "step": 4033
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5368759632110596,
+      "learning_rate": 4.893629394660323e-05,
+      "loss": 1.4156,
+      "step": 4034
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5667952299118042,
+      "learning_rate": 4.89360296061327e-05,
+      "loss": 1.4689,
+      "step": 4035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6296359300613403,
+      "learning_rate": 4.893576526566218e-05,
+      "loss": 1.3854,
+      "step": 4036
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.796862840652466,
+      "learning_rate": 4.893550092519165e-05,
+      "loss": 1.5151,
+      "step": 4037
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.892205834388733,
+      "learning_rate": 4.8935236584721125e-05,
+      "loss": 1.4856,
+      "step": 4038
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7873408794403076,
+      "learning_rate": 4.89349722442506e-05,
+      "loss": 1.4281,
+      "step": 4039
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.586806058883667,
+      "learning_rate": 4.893470790378007e-05,
+      "loss": 1.6625,
+      "step": 4040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4870221614837646,
+      "learning_rate": 4.8934443563309546e-05,
+      "loss": 1.5608,
+      "step": 4041
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3313870429992676,
+      "learning_rate": 4.893417922283902e-05,
+      "loss": 1.4418,
+      "step": 4042
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0895862579345703,
+      "learning_rate": 4.893391488236849e-05,
+      "loss": 1.4518,
+      "step": 4043
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9638944864273071,
+      "learning_rate": 4.8933650541897966e-05,
+      "loss": 1.429,
+      "step": 4044
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6591532230377197,
+      "learning_rate": 4.893338620142744e-05,
+      "loss": 1.3647,
+      "step": 4045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4138423204421997,
+      "learning_rate": 4.8933121860956914e-05,
+      "loss": 1.3475,
+      "step": 4046
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1151750087738037,
+      "learning_rate": 4.8932857520486394e-05,
+      "loss": 1.4708,
+      "step": 4047
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4639825820922852,
+      "learning_rate": 4.893259318001586e-05,
+      "loss": 1.3858,
+      "step": 4048
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4567307233810425,
+      "learning_rate": 4.893232883954534e-05,
+      "loss": 1.3833,
+      "step": 4049
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.507777214050293,
+      "learning_rate": 4.893206449907481e-05,
+      "loss": 1.3684,
+      "step": 4050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7357571125030518,
+      "learning_rate": 4.893180015860429e-05,
+      "loss": 1.4462,
+      "step": 4051
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4838752746582031,
+      "learning_rate": 4.893153581813376e-05,
+      "loss": 2.1195,
+      "step": 4052
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8640333414077759,
+      "learning_rate": 4.8931271477663235e-05,
+      "loss": 1.6138,
+      "step": 4053
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4314044713974,
+      "learning_rate": 4.893100713719271e-05,
+      "loss": 1.4548,
+      "step": 4054
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7563830614089966,
+      "learning_rate": 4.8930742796722176e-05,
+      "loss": 1.438,
+      "step": 4055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8449575901031494,
+      "learning_rate": 4.8930478456251656e-05,
+      "loss": 1.4848,
+      "step": 4056
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7384312152862549,
+      "learning_rate": 4.893021411578113e-05,
+      "loss": 1.3173,
+      "step": 4057
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7245230674743652,
+      "learning_rate": 4.89299497753106e-05,
+      "loss": 1.3034,
+      "step": 4058
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.959867238998413,
+      "learning_rate": 4.892968543484008e-05,
+      "loss": 1.3445,
+      "step": 4059
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4871667623519897,
+      "learning_rate": 4.892942109436955e-05,
+      "loss": 1.3072,
+      "step": 4060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6499559879302979,
+      "learning_rate": 4.8929156753899024e-05,
+      "loss": 1.4191,
+      "step": 4061
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.943636178970337,
+      "learning_rate": 4.8928892413428504e-05,
+      "loss": 1.4373,
+      "step": 4062
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.578331470489502,
+      "learning_rate": 4.892862807295797e-05,
+      "loss": 1.3751,
+      "step": 4063
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4599201679229736,
+      "learning_rate": 4.8928363732487445e-05,
+      "loss": 1.3334,
+      "step": 4064
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0316829681396484,
+      "learning_rate": 4.892809939201692e-05,
+      "loss": 1.282,
+      "step": 4065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7275428771972656,
+      "learning_rate": 4.892783505154639e-05,
+      "loss": 1.3791,
+      "step": 4066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8475077152252197,
+      "learning_rate": 4.892757071107587e-05,
+      "loss": 1.4066,
+      "step": 4067
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9711512327194214,
+      "learning_rate": 4.892730637060534e-05,
+      "loss": 1.5161,
+      "step": 4068
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7988444566726685,
+      "learning_rate": 4.892704203013482e-05,
+      "loss": 1.4922,
+      "step": 4069
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5905561447143555,
+      "learning_rate": 4.8926777689664286e-05,
+      "loss": 1.7109,
+      "step": 4070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.947694182395935,
+      "learning_rate": 4.8926513349193766e-05,
+      "loss": 1.5441,
+      "step": 4071
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2742528915405273,
+      "learning_rate": 4.892624900872324e-05,
+      "loss": 1.4636,
+      "step": 4072
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9298200607299805,
+      "learning_rate": 4.8925984668252713e-05,
+      "loss": 1.4449,
+      "step": 4073
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.494469404220581,
+      "learning_rate": 4.892572032778219e-05,
+      "loss": 1.7961,
+      "step": 4074
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8519567251205444,
+      "learning_rate": 4.892545598731166e-05,
+      "loss": 1.443,
+      "step": 4075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5510210990905762,
+      "learning_rate": 4.8925191646841134e-05,
+      "loss": 1.421,
+      "step": 4076
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7602894306182861,
+      "learning_rate": 4.892492730637061e-05,
+      "loss": 1.5289,
+      "step": 4077
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4582189321517944,
+      "learning_rate": 4.892466296590008e-05,
+      "loss": 1.3393,
+      "step": 4078
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6140848398208618,
+      "learning_rate": 4.8924398625429555e-05,
+      "loss": 1.3274,
+      "step": 4079
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5689442157745361,
+      "learning_rate": 4.892413428495903e-05,
+      "loss": 1.3523,
+      "step": 4080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7335442304611206,
+      "learning_rate": 4.89238699444885e-05,
+      "loss": 1.4622,
+      "step": 4081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.469595193862915,
+      "learning_rate": 4.892360560401798e-05,
+      "loss": 1.4845,
+      "step": 4082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8090909719467163,
+      "learning_rate": 4.892334126354745e-05,
+      "loss": 1.4066,
+      "step": 4083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5362920761108398,
+      "learning_rate": 4.892307692307693e-05,
+      "loss": 1.3354,
+      "step": 4084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.238858699798584,
+      "learning_rate": 4.8922812582606396e-05,
+      "loss": 1.5748,
+      "step": 4085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.657525897026062,
+      "learning_rate": 4.892254824213588e-05,
+      "loss": 1.5253,
+      "step": 4086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4789751768112183,
+      "learning_rate": 4.892228390166535e-05,
+      "loss": 1.3932,
+      "step": 4087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6216344833374023,
+      "learning_rate": 4.8922019561194824e-05,
+      "loss": 1.4521,
+      "step": 4088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4783029556274414,
+      "learning_rate": 4.89217552207243e-05,
+      "loss": 1.3191,
+      "step": 4089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7028850317001343,
+      "learning_rate": 4.8921490880253764e-05,
+      "loss": 1.3147,
+      "step": 4090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7613414525985718,
+      "learning_rate": 4.8921226539783245e-05,
+      "loss": 1.3563,
+      "step": 4091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6152375936508179,
+      "learning_rate": 4.892096219931272e-05,
+      "loss": 1.4146,
+      "step": 4092
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1753180027008057,
+      "learning_rate": 4.892069785884219e-05,
+      "loss": 1.6154,
+      "step": 4093
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1136484146118164,
+      "learning_rate": 4.8920433518371665e-05,
+      "loss": 1.4171,
+      "step": 4094
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9763586521148682,
+      "learning_rate": 4.892016917790114e-05,
+      "loss": 1.3419,
+      "step": 4095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.244645357131958,
+      "learning_rate": 4.891990483743061e-05,
+      "loss": 1.805,
+      "step": 4096
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6777331829071045,
+      "learning_rate": 4.891964049696009e-05,
+      "loss": 1.3094,
+      "step": 4097
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6226330995559692,
+      "learning_rate": 4.891937615648956e-05,
+      "loss": 1.283,
+      "step": 4098
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.735393762588501,
+      "learning_rate": 4.891911181601903e-05,
+      "loss": 1.3552,
+      "step": 4099
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.616206407546997,
+      "learning_rate": 4.891884747554851e-05,
+      "loss": 1.4021,
+      "step": 4100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.706624150276184,
+      "learning_rate": 4.891858313507798e-05,
+      "loss": 1.4715,
+      "step": 4101
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8571603298187256,
+      "learning_rate": 4.891831879460746e-05,
+      "loss": 1.4221,
+      "step": 4102
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6840075254440308,
+      "learning_rate": 4.891805445413693e-05,
+      "loss": 1.2692,
+      "step": 4103
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6806528568267822,
+      "learning_rate": 4.891779011366641e-05,
+      "loss": 1.4461,
+      "step": 4104
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.536440134048462,
+      "learning_rate": 4.8917525773195875e-05,
+      "loss": 1.4,
+      "step": 4105
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5399515628814697,
+      "learning_rate": 4.8917261432725355e-05,
+      "loss": 1.3787,
+      "step": 4106
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8323187828063965,
+      "learning_rate": 4.891699709225483e-05,
+      "loss": 2.0684,
+      "step": 4107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.973846673965454,
+      "learning_rate": 4.89167327517843e-05,
+      "loss": 1.401,
+      "step": 4108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.757946729660034,
+      "learning_rate": 4.8916468411313776e-05,
+      "loss": 1.6697,
+      "step": 4109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9129546880722046,
+      "learning_rate": 4.891620407084325e-05,
+      "loss": 1.5256,
+      "step": 4110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0673606395721436,
+      "learning_rate": 4.891593973037272e-05,
+      "loss": 1.4597,
+      "step": 4111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.67918860912323,
+      "learning_rate": 4.8915675389902196e-05,
+      "loss": 1.2904,
+      "step": 4112
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.023698568344116,
+      "learning_rate": 4.891541104943167e-05,
+      "loss": 1.4058,
+      "step": 4113
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0588762760162354,
+      "learning_rate": 4.8915146708961143e-05,
+      "loss": 1.4922,
+      "step": 4114
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7008336782455444,
+      "learning_rate": 4.891488236849062e-05,
+      "loss": 1.307,
+      "step": 4115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9467004537582397,
+      "learning_rate": 4.891461802802009e-05,
+      "loss": 1.4336,
+      "step": 4116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0842816829681396,
+      "learning_rate": 4.891435368754957e-05,
+      "loss": 1.3415,
+      "step": 4117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8049156665802002,
+      "learning_rate": 4.891408934707904e-05,
+      "loss": 1.4802,
+      "step": 4118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.286695957183838,
+      "learning_rate": 4.891382500660852e-05,
+      "loss": 1.4966,
+      "step": 4119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6848294734954834,
+      "learning_rate": 4.8913560666137985e-05,
+      "loss": 1.4426,
+      "step": 4120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2180020809173584,
+      "learning_rate": 4.8913296325667465e-05,
+      "loss": 1.5378,
+      "step": 4121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0756568908691406,
+      "learning_rate": 4.891303198519694e-05,
+      "loss": 1.5406,
+      "step": 4122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4967845678329468,
+      "learning_rate": 4.891276764472641e-05,
+      "loss": 1.306,
+      "step": 4123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6412100791931152,
+      "learning_rate": 4.8912503304255886e-05,
+      "loss": 1.3298,
+      "step": 4124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8231801986694336,
+      "learning_rate": 4.891223896378536e-05,
+      "loss": 1.2495,
+      "step": 4125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9115253686904907,
+      "learning_rate": 4.891197462331483e-05,
+      "loss": 1.3454,
+      "step": 4126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6312052011489868,
+      "learning_rate": 4.891171028284431e-05,
+      "loss": 1.3257,
+      "step": 4127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7169677019119263,
+      "learning_rate": 4.891144594237378e-05,
+      "loss": 1.3411,
+      "step": 4128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1529746055603027,
+      "learning_rate": 4.8911181601903254e-05,
+      "loss": 1.5557,
+      "step": 4129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6368144750595093,
+      "learning_rate": 4.891091726143273e-05,
+      "loss": 1.4967,
+      "step": 4130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6481378078460693,
+      "learning_rate": 4.89106529209622e-05,
+      "loss": 1.4476,
+      "step": 4131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.598250389099121,
+      "learning_rate": 4.891038858049168e-05,
+      "loss": 1.6899,
+      "step": 4132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3749133348464966,
+      "learning_rate": 4.891012424002115e-05,
+      "loss": 1.3927,
+      "step": 4133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6257290840148926,
+      "learning_rate": 4.890985989955063e-05,
+      "loss": 1.4846,
+      "step": 4134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0308563709259033,
+      "learning_rate": 4.8909595559080095e-05,
+      "loss": 1.4081,
+      "step": 4135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8451356887817383,
+      "learning_rate": 4.890933121860957e-05,
+      "loss": 1.3809,
+      "step": 4136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7176971435546875,
+      "learning_rate": 4.890906687813905e-05,
+      "loss": 1.368,
+      "step": 4137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.757267713546753,
+      "learning_rate": 4.8908802537668516e-05,
+      "loss": 1.457,
+      "step": 4138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3560264110565186,
+      "learning_rate": 4.8908538197197996e-05,
+      "loss": 1.5599,
+      "step": 4139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7746295928955078,
+      "learning_rate": 4.890827385672746e-05,
+      "loss": 1.3997,
+      "step": 4140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.748105764389038,
+      "learning_rate": 4.8908009516256943e-05,
+      "loss": 1.3321,
+      "step": 4141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8973727226257324,
+      "learning_rate": 4.890774517578642e-05,
+      "loss": 1.4184,
+      "step": 4142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.413196563720703,
+      "learning_rate": 4.890748083531589e-05,
+      "loss": 1.4149,
+      "step": 4143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8644026517868042,
+      "learning_rate": 4.8907216494845364e-05,
+      "loss": 1.5135,
+      "step": 4144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3766674995422363,
+      "learning_rate": 4.890695215437484e-05,
+      "loss": 1.3725,
+      "step": 4145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.682594656944275,
+      "learning_rate": 4.890668781390431e-05,
+      "loss": 1.4297,
+      "step": 4146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0762226581573486,
+      "learning_rate": 4.8906423473433785e-05,
+      "loss": 1.4842,
+      "step": 4147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5053412914276123,
+      "learning_rate": 4.890615913296326e-05,
+      "loss": 1.3422,
+      "step": 4148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6945605278015137,
+      "learning_rate": 4.890589479249273e-05,
+      "loss": 2.3209,
+      "step": 4149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.481375217437744,
+      "learning_rate": 4.8905630452022206e-05,
+      "loss": 1.5407,
+      "step": 4150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2659552097320557,
+      "learning_rate": 4.890536611155168e-05,
+      "loss": 1.4943,
+      "step": 4151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6153688430786133,
+      "learning_rate": 4.890510177108115e-05,
+      "loss": 1.5156,
+      "step": 4152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1600534915924072,
+      "learning_rate": 4.8904837430610626e-05,
+      "loss": 1.3974,
+      "step": 4153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6069055795669556,
+      "learning_rate": 4.8904573090140107e-05,
+      "loss": 1.3276,
+      "step": 4154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.61143958568573,
+      "learning_rate": 4.8904308749669573e-05,
+      "loss": 1.3669,
+      "step": 4155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2710585594177246,
+      "learning_rate": 4.8904044409199054e-05,
+      "loss": 1.5582,
+      "step": 4156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9710814952850342,
+      "learning_rate": 4.890378006872852e-05,
+      "loss": 1.5635,
+      "step": 4157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8545689582824707,
+      "learning_rate": 4.8903515728258e-05,
+      "loss": 1.2516,
+      "step": 4158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.892712950706482,
+      "learning_rate": 4.8903251387787474e-05,
+      "loss": 1.6009,
+      "step": 4159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4130215644836426,
+      "learning_rate": 4.890298704731695e-05,
+      "loss": 1.4168,
+      "step": 4160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4407331943511963,
+      "learning_rate": 4.890272270684642e-05,
+      "loss": 1.5238,
+      "step": 4161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0080983638763428,
+      "learning_rate": 4.890245836637589e-05,
+      "loss": 1.4419,
+      "step": 4162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5240812301635742,
+      "learning_rate": 4.890219402590537e-05,
+      "loss": 1.3486,
+      "step": 4163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.427043080329895,
+      "learning_rate": 4.890192968543484e-05,
+      "loss": 1.3414,
+      "step": 4164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.440526008605957,
+      "learning_rate": 4.8901665344964316e-05,
+      "loss": 1.5968,
+      "step": 4165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5902023315429688,
+      "learning_rate": 4.890140100449379e-05,
+      "loss": 1.5202,
+      "step": 4166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.52104651927948,
+      "learning_rate": 4.890113666402326e-05,
+      "loss": 1.377,
+      "step": 4167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6128793954849243,
+      "learning_rate": 4.8900872323552737e-05,
+      "loss": 1.3043,
+      "step": 4168
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6549023389816284,
+      "learning_rate": 4.890060798308222e-05,
+      "loss": 1.453,
+      "step": 4169
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7931101322174072,
+      "learning_rate": 4.8900343642611684e-05,
+      "loss": 1.2774,
+      "step": 4170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.031888484954834,
+      "learning_rate": 4.8900079302141164e-05,
+      "loss": 1.5673,
+      "step": 4171
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7893946170806885,
+      "learning_rate": 4.889981496167063e-05,
+      "loss": 1.3023,
+      "step": 4172
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.03775954246521,
+      "learning_rate": 4.8899550621200104e-05,
+      "loss": 1.4189,
+      "step": 4173
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2449135780334473,
+      "learning_rate": 4.8899286280729585e-05,
+      "loss": 1.2113,
+      "step": 4174
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5419014692306519,
+      "learning_rate": 4.889902194025905e-05,
+      "loss": 2.2103,
+      "step": 4175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8375002145767212,
+      "learning_rate": 4.889875759978853e-05,
+      "loss": 1.5166,
+      "step": 4176
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.772574543952942,
+      "learning_rate": 4.8898493259318e-05,
+      "loss": 1.3539,
+      "step": 4177
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0986950397491455,
+      "learning_rate": 4.889822891884748e-05,
+      "loss": 1.4035,
+      "step": 4178
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.757217288017273,
+      "learning_rate": 4.889796457837695e-05,
+      "loss": 1.4312,
+      "step": 4179
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6178547143936157,
+      "learning_rate": 4.8897700237906426e-05,
+      "loss": 1.4621,
+      "step": 4180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1565396785736084,
+      "learning_rate": 4.88974358974359e-05,
+      "loss": 1.6771,
+      "step": 4181
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.140214443206787,
+      "learning_rate": 4.889717155696537e-05,
+      "loss": 1.5686,
+      "step": 4182
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.009404182434082,
+      "learning_rate": 4.889690721649485e-05,
+      "loss": 1.4491,
+      "step": 4183
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.057548999786377,
+      "learning_rate": 4.889664287602432e-05,
+      "loss": 1.6036,
+      "step": 4184
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2076785564422607,
+      "learning_rate": 4.8896378535553794e-05,
+      "loss": 1.3348,
+      "step": 4185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0329172611236572,
+      "learning_rate": 4.889611419508327e-05,
+      "loss": 1.4626,
+      "step": 4186
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.466107964515686,
+      "learning_rate": 4.889584985461274e-05,
+      "loss": 1.3259,
+      "step": 4187
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.881234884262085,
+      "learning_rate": 4.8895585514142215e-05,
+      "loss": 1.3084,
+      "step": 4188
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9611680507659912,
+      "learning_rate": 4.8895321173671695e-05,
+      "loss": 1.359,
+      "step": 4189
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.445009708404541,
+      "learning_rate": 4.889505683320116e-05,
+      "loss": 1.4026,
+      "step": 4190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5791802406311035,
+      "learning_rate": 4.889479249273064e-05,
+      "loss": 1.2895,
+      "step": 4191
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.526733636856079,
+      "learning_rate": 4.889452815226011e-05,
+      "loss": 1.3937,
+      "step": 4192
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6770235300064087,
+      "learning_rate": 4.889426381178959e-05,
+      "loss": 1.5095,
+      "step": 4193
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.190314531326294,
+      "learning_rate": 4.889399947131906e-05,
+      "loss": 1.7416,
+      "step": 4194
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7292609214782715,
+      "learning_rate": 4.8893735130848537e-05,
+      "loss": 1.316,
+      "step": 4195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9626076221466064,
+      "learning_rate": 4.889347079037801e-05,
+      "loss": 1.6364,
+      "step": 4196
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1480307579040527,
+      "learning_rate": 4.8893206449907484e-05,
+      "loss": 1.5109,
+      "step": 4197
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4754117727279663,
+      "learning_rate": 4.889294210943696e-05,
+      "loss": 1.4934,
+      "step": 4198
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5615631341934204,
+      "learning_rate": 4.889267776896643e-05,
+      "loss": 1.3248,
+      "step": 4199
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.636641502380371,
+      "learning_rate": 4.8892413428495904e-05,
+      "loss": 1.3899,
+      "step": 4200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8054293394088745,
+      "learning_rate": 4.889214908802538e-05,
+      "loss": 1.5985,
+      "step": 4201
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6690672636032104,
+      "learning_rate": 4.889188474755485e-05,
+      "loss": 1.4378,
+      "step": 4202
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7867765426635742,
+      "learning_rate": 4.8891620407084325e-05,
+      "loss": 1.4716,
+      "step": 4203
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6577727794647217,
+      "learning_rate": 4.8891356066613805e-05,
+      "loss": 1.2601,
+      "step": 4204
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8101943731307983,
+      "learning_rate": 4.889109172614327e-05,
+      "loss": 1.5286,
+      "step": 4205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.038252830505371,
+      "learning_rate": 4.889082738567275e-05,
+      "loss": 1.5685,
+      "step": 4206
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5333331823349,
+      "learning_rate": 4.889056304520222e-05,
+      "loss": 1.352,
+      "step": 4207
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.575819730758667,
+      "learning_rate": 4.889029870473169e-05,
+      "loss": 1.3776,
+      "step": 4208
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1157251596450806,
+      "learning_rate": 4.889003436426117e-05,
+      "loss": 2.3509,
+      "step": 4209
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6279090642929077,
+      "learning_rate": 4.888977002379064e-05,
+      "loss": 1.8359,
+      "step": 4210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6587222814559937,
+      "learning_rate": 4.888950568332012e-05,
+      "loss": 1.3559,
+      "step": 4211
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7206817865371704,
+      "learning_rate": 4.888924134284959e-05,
+      "loss": 1.4218,
+      "step": 4212
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6115144491195679,
+      "learning_rate": 4.888897700237907e-05,
+      "loss": 1.3573,
+      "step": 4213
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.888434886932373,
+      "learning_rate": 4.888871266190854e-05,
+      "loss": 1.6615,
+      "step": 4214
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9320800304412842,
+      "learning_rate": 4.8888448321438015e-05,
+      "loss": 1.4842,
+      "step": 4215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.820798397064209,
+      "learning_rate": 4.888818398096749e-05,
+      "loss": 1.4324,
+      "step": 4216
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.191680669784546,
+      "learning_rate": 4.888791964049696e-05,
+      "loss": 1.3946,
+      "step": 4217
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.883582830429077,
+      "learning_rate": 4.8887655300026435e-05,
+      "loss": 1.4884,
+      "step": 4218
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.532494068145752,
+      "learning_rate": 4.888739095955591e-05,
+      "loss": 1.4954,
+      "step": 4219
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.935547113418579,
+      "learning_rate": 4.888712661908538e-05,
+      "loss": 1.4593,
+      "step": 4220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9101383686065674,
+      "learning_rate": 4.8886862278614856e-05,
+      "loss": 1.467,
+      "step": 4221
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5748515129089355,
+      "learning_rate": 4.888659793814433e-05,
+      "loss": 1.364,
+      "step": 4222
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6298167705535889,
+      "learning_rate": 4.88863335976738e-05,
+      "loss": 1.3821,
+      "step": 4223
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4715282917022705,
+      "learning_rate": 4.8886069257203284e-05,
+      "loss": 1.7334,
+      "step": 4224
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4998879432678223,
+      "learning_rate": 4.888580491673275e-05,
+      "loss": 1.4381,
+      "step": 4225
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.625022530555725,
+      "learning_rate": 4.888554057626223e-05,
+      "loss": 1.4128,
+      "step": 4226
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1034152507781982,
+      "learning_rate": 4.88852762357917e-05,
+      "loss": 1.5139,
+      "step": 4227
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.822495937347412,
+      "learning_rate": 4.888501189532118e-05,
+      "loss": 1.4813,
+      "step": 4228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8994834423065186,
+      "learning_rate": 4.888474755485065e-05,
+      "loss": 1.4966,
+      "step": 4229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.467714786529541,
+      "learning_rate": 4.8884483214380125e-05,
+      "loss": 1.3444,
+      "step": 4230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.061236619949341,
+      "learning_rate": 4.88842188739096e-05,
+      "loss": 1.4721,
+      "step": 4231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5612519979476929,
+      "learning_rate": 4.888395453343907e-05,
+      "loss": 1.4033,
+      "step": 4232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1456077098846436,
+      "learning_rate": 4.8883690192968546e-05,
+      "loss": 1.4741,
+      "step": 4233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5568621158599854,
+      "learning_rate": 4.888342585249802e-05,
+      "loss": 1.5444,
+      "step": 4234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.889385461807251,
+      "learning_rate": 4.888316151202749e-05,
+      "loss": 1.8532,
+      "step": 4235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.837411642074585,
+      "learning_rate": 4.8882897171556967e-05,
+      "loss": 1.3523,
+      "step": 4236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6838088035583496,
+      "learning_rate": 4.888263283108644e-05,
+      "loss": 1.2291,
+      "step": 4237
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.238971710205078,
+      "learning_rate": 4.8882368490615914e-05,
+      "loss": 1.4068,
+      "step": 4238
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9410706758499146,
+      "learning_rate": 4.8882104150145394e-05,
+      "loss": 1.3799,
+      "step": 4239
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7997370958328247,
+      "learning_rate": 4.888183980967486e-05,
+      "loss": 1.567,
+      "step": 4240
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.096099615097046,
+      "learning_rate": 4.888157546920434e-05,
+      "loss": 1.4227,
+      "step": 4241
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.281593680381775,
+      "learning_rate": 4.888131112873381e-05,
+      "loss": 2.0517,
+      "step": 4242
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.10709285736084,
+      "learning_rate": 4.888104678826329e-05,
+      "loss": 1.5835,
+      "step": 4243
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.570536732673645,
+      "learning_rate": 4.888078244779276e-05,
+      "loss": 1.3675,
+      "step": 4244
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.323246717453003,
+      "learning_rate": 4.888051810732223e-05,
+      "loss": 1.4271,
+      "step": 4245
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.802712321281433,
+      "learning_rate": 4.888025376685171e-05,
+      "loss": 2.6667,
+      "step": 4246
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.887972354888916,
+      "learning_rate": 4.8879989426381176e-05,
+      "loss": 1.5089,
+      "step": 4247
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6502374410629272,
+      "learning_rate": 4.8879725085910656e-05,
+      "loss": 1.4253,
+      "step": 4248
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7204265594482422,
+      "learning_rate": 4.887946074544013e-05,
+      "loss": 1.46,
+      "step": 4249
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7040371894836426,
+      "learning_rate": 4.88791964049696e-05,
+      "loss": 1.3427,
+      "step": 4250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6461181640625,
+      "learning_rate": 4.887893206449908e-05,
+      "loss": 1.5337,
+      "step": 4251
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.744086503982544,
+      "learning_rate": 4.887866772402855e-05,
+      "loss": 1.4984,
+      "step": 4252
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1960840225219727,
+      "learning_rate": 4.8878403383558024e-05,
+      "loss": 1.661,
+      "step": 4253
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5967321395874023,
+      "learning_rate": 4.88781390430875e-05,
+      "loss": 1.3631,
+      "step": 4254
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.501006841659546,
+      "learning_rate": 4.887787470261697e-05,
+      "loss": 1.3024,
+      "step": 4255
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9070687294006348,
+      "learning_rate": 4.8877610362146445e-05,
+      "loss": 1.3584,
+      "step": 4256
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3393137454986572,
+      "learning_rate": 4.887734602167592e-05,
+      "loss": 1.2275,
+      "step": 4257
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6986409425735474,
+      "learning_rate": 4.887708168120539e-05,
+      "loss": 1.4999,
+      "step": 4258
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3083146810531616,
+      "learning_rate": 4.887681734073487e-05,
+      "loss": 1.2987,
+      "step": 4259
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2272369861602783,
+      "learning_rate": 4.887655300026434e-05,
+      "loss": 1.4489,
+      "step": 4260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0199408531188965,
+      "learning_rate": 4.887628865979382e-05,
+      "loss": 1.4581,
+      "step": 4261
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6431366205215454,
+      "learning_rate": 4.8876024319323286e-05,
+      "loss": 1.3878,
+      "step": 4262
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.780446171760559,
+      "learning_rate": 4.8875759978852766e-05,
+      "loss": 1.4973,
+      "step": 4263
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7655489444732666,
+      "learning_rate": 4.887549563838224e-05,
+      "loss": 1.4624,
+      "step": 4264
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3601247072219849,
+      "learning_rate": 4.8875231297911714e-05,
+      "loss": 1.368,
+      "step": 4265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5946438312530518,
+      "learning_rate": 4.887496695744119e-05,
+      "loss": 1.4766,
+      "step": 4266
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0409679412841797,
+      "learning_rate": 4.887470261697066e-05,
+      "loss": 1.5294,
+      "step": 4267
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6858007907867432,
+      "learning_rate": 4.8874438276500134e-05,
+      "loss": 1.2456,
+      "step": 4268
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6750404834747314,
+      "learning_rate": 4.887417393602961e-05,
+      "loss": 1.4735,
+      "step": 4269
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6991618871688843,
+      "learning_rate": 4.887390959555908e-05,
+      "loss": 1.424,
+      "step": 4270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8806718587875366,
+      "learning_rate": 4.8873645255088555e-05,
+      "loss": 1.5423,
+      "step": 4271
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3339468240737915,
+      "learning_rate": 4.887338091461803e-05,
+      "loss": 1.1841,
+      "step": 4272
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.325002670288086,
+      "learning_rate": 4.88731165741475e-05,
+      "loss": 1.2785,
+      "step": 4273
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8322325944900513,
+      "learning_rate": 4.887285223367698e-05,
+      "loss": 1.5373,
+      "step": 4274
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6962506771087646,
+      "learning_rate": 4.887258789320645e-05,
+      "loss": 1.3773,
+      "step": 4275
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4705233573913574,
+      "learning_rate": 4.887232355273593e-05,
+      "loss": 1.2447,
+      "step": 4276
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9023220539093018,
+      "learning_rate": 4.8872059212265396e-05,
+      "loss": 2.2886,
+      "step": 4277
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4329520463943481,
+      "learning_rate": 4.887179487179488e-05,
+      "loss": 1.3629,
+      "step": 4278
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6197513341903687,
+      "learning_rate": 4.887153053132435e-05,
+      "loss": 1.3896,
+      "step": 4279
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9331204891204834,
+      "learning_rate": 4.887126619085382e-05,
+      "loss": 1.4593,
+      "step": 4280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4507783651351929,
+      "learning_rate": 4.88710018503833e-05,
+      "loss": 2.3884,
+      "step": 4281
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9858003854751587,
+      "learning_rate": 4.8870737509912764e-05,
+      "loss": 1.4205,
+      "step": 4282
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7892544269561768,
+      "learning_rate": 4.8870473169442245e-05,
+      "loss": 1.2864,
+      "step": 4283
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.505829334259033,
+      "learning_rate": 4.887020882897172e-05,
+      "loss": 1.7688,
+      "step": 4284
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.57625150680542,
+      "learning_rate": 4.886994448850119e-05,
+      "loss": 1.4456,
+      "step": 4285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3955470323562622,
+      "learning_rate": 4.8869680148030665e-05,
+      "loss": 2.1217,
+      "step": 4286
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4368001222610474,
+      "learning_rate": 4.886941580756014e-05,
+      "loss": 2.4349,
+      "step": 4287
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9139419794082642,
+      "learning_rate": 4.886915146708961e-05,
+      "loss": 1.4964,
+      "step": 4288
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9912919998168945,
+      "learning_rate": 4.886888712661909e-05,
+      "loss": 1.5122,
+      "step": 4289
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.824307918548584,
+      "learning_rate": 4.886862278614856e-05,
+      "loss": 1.6379,
+      "step": 4290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7413376569747925,
+      "learning_rate": 4.886835844567803e-05,
+      "loss": 1.4015,
+      "step": 4291
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8660892248153687,
+      "learning_rate": 4.886809410520751e-05,
+      "loss": 1.3962,
+      "step": 4292
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5669724941253662,
+      "learning_rate": 4.886782976473698e-05,
+      "loss": 1.4157,
+      "step": 4293
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3434226512908936,
+      "learning_rate": 4.886756542426646e-05,
+      "loss": 1.3798,
+      "step": 4294
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8487166166305542,
+      "learning_rate": 4.886730108379593e-05,
+      "loss": 1.4431,
+      "step": 4295
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7434356212615967,
+      "learning_rate": 4.886703674332541e-05,
+      "loss": 1.5975,
+      "step": 4296
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.028651714324951,
+      "learning_rate": 4.8866772402854875e-05,
+      "loss": 1.561,
+      "step": 4297
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5504354238510132,
+      "learning_rate": 4.8866508062384355e-05,
+      "loss": 1.9472,
+      "step": 4298
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6598937511444092,
+      "learning_rate": 4.886624372191383e-05,
+      "loss": 2.6479,
+      "step": 4299
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6879847049713135,
+      "learning_rate": 4.88659793814433e-05,
+      "loss": 1.4258,
+      "step": 4300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.011641025543213,
+      "learning_rate": 4.8865715040972776e-05,
+      "loss": 1.3867,
+      "step": 4301
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9162224531173706,
+      "learning_rate": 4.886545070050225e-05,
+      "loss": 1.5721,
+      "step": 4302
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8014309406280518,
+      "learning_rate": 4.886518636003172e-05,
+      "loss": 2.4012,
+      "step": 4303
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9317352771759033,
+      "learning_rate": 4.8864922019561196e-05,
+      "loss": 1.4771,
+      "step": 4304
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6897631883621216,
+      "learning_rate": 4.886465767909067e-05,
+      "loss": 1.3936,
+      "step": 4305
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.538857340812683,
+      "learning_rate": 4.8864393338620144e-05,
+      "loss": 1.48,
+      "step": 4306
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.925352692604065,
+      "learning_rate": 4.886412899814962e-05,
+      "loss": 1.5613,
+      "step": 4307
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.110205888748169,
+      "learning_rate": 4.886386465767909e-05,
+      "loss": 1.4026,
+      "step": 4308
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5099902153015137,
+      "learning_rate": 4.886360031720857e-05,
+      "loss": 1.3058,
+      "step": 4309
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6514545679092407,
+      "learning_rate": 4.886333597673804e-05,
+      "loss": 1.2029,
+      "step": 4310
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9870061874389648,
+      "learning_rate": 4.886307163626752e-05,
+      "loss": 1.4775,
+      "step": 4311
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.665409803390503,
+      "learning_rate": 4.8862807295796985e-05,
+      "loss": 1.4609,
+      "step": 4312
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6766489744186401,
+      "learning_rate": 4.8862542955326465e-05,
+      "loss": 1.3874,
+      "step": 4313
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2869508266448975,
+      "learning_rate": 4.886227861485594e-05,
+      "loss": 1.4996,
+      "step": 4314
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0901589393615723,
+      "learning_rate": 4.886201427438541e-05,
+      "loss": 1.6941,
+      "step": 4315
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6659059524536133,
+      "learning_rate": 4.8861749933914886e-05,
+      "loss": 1.4844,
+      "step": 4316
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7992974519729614,
+      "learning_rate": 4.886148559344435e-05,
+      "loss": 1.3348,
+      "step": 4317
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8821656703948975,
+      "learning_rate": 4.886122125297383e-05,
+      "loss": 1.4541,
+      "step": 4318
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8540982007980347,
+      "learning_rate": 4.886095691250331e-05,
+      "loss": 1.575,
+      "step": 4319
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0162906646728516,
+      "learning_rate": 4.886069257203278e-05,
+      "loss": 1.5126,
+      "step": 4320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.575574278831482,
+      "learning_rate": 4.8860428231562254e-05,
+      "loss": 1.2998,
+      "step": 4321
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3982555866241455,
+      "learning_rate": 4.886016389109173e-05,
+      "loss": 1.4766,
+      "step": 4322
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0106618404388428,
+      "learning_rate": 4.88598995506212e-05,
+      "loss": 1.527,
+      "step": 4323
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.796189308166504,
+      "learning_rate": 4.885963521015068e-05,
+      "loss": 1.523,
+      "step": 4324
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5246219635009766,
+      "learning_rate": 4.885937086968015e-05,
+      "loss": 1.2912,
+      "step": 4325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.578019142150879,
+      "learning_rate": 4.885910652920962e-05,
+      "loss": 1.4281,
+      "step": 4326
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8435851335525513,
+      "learning_rate": 4.8858842188739095e-05,
+      "loss": 1.5717,
+      "step": 4327
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6849327087402344,
+      "learning_rate": 4.885857784826857e-05,
+      "loss": 2.7357,
+      "step": 4328
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7271984815597534,
+      "learning_rate": 4.885831350779805e-05,
+      "loss": 1.5491,
+      "step": 4329
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.64630389213562,
+      "learning_rate": 4.8858049167327516e-05,
+      "loss": 1.4948,
+      "step": 4330
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2714166641235352,
+      "learning_rate": 4.8857784826856996e-05,
+      "loss": 1.1953,
+      "step": 4331
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7172943353652954,
+      "learning_rate": 4.885752048638646e-05,
+      "loss": 1.5268,
+      "step": 4332
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4972460269927979,
+      "learning_rate": 4.8857256145915944e-05,
+      "loss": 1.0737,
+      "step": 4333
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5079245567321777,
+      "learning_rate": 4.885699180544542e-05,
+      "loss": 1.247,
+      "step": 4334
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3691178560256958,
+      "learning_rate": 4.885672746497489e-05,
+      "loss": 1.3325,
+      "step": 4335
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.622091770172119,
+      "learning_rate": 4.8856463124504364e-05,
+      "loss": 1.6924,
+      "step": 4336
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6732864379882812,
+      "learning_rate": 4.885619878403384e-05,
+      "loss": 1.3092,
+      "step": 4337
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3578975200653076,
+      "learning_rate": 4.885593444356331e-05,
+      "loss": 1.2621,
+      "step": 4338
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.187931776046753,
+      "learning_rate": 4.8855670103092785e-05,
+      "loss": 1.4419,
+      "step": 4339
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9420828819274902,
+      "learning_rate": 4.885540576262226e-05,
+      "loss": 1.3882,
+      "step": 4340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4336390495300293,
+      "learning_rate": 4.885514142215173e-05,
+      "loss": 1.3676,
+      "step": 4341
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3494127988815308,
+      "learning_rate": 4.8854877081681206e-05,
+      "loss": 1.3949,
+      "step": 4342
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7557060718536377,
+      "learning_rate": 4.885461274121068e-05,
+      "loss": 1.3267,
+      "step": 4343
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5077762603759766,
+      "learning_rate": 4.885434840074016e-05,
+      "loss": 1.322,
+      "step": 4344
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.793540358543396,
+      "learning_rate": 4.8854084060269626e-05,
+      "loss": 1.3727,
+      "step": 4345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7326209545135498,
+      "learning_rate": 4.885381971979911e-05,
+      "loss": 1.5012,
+      "step": 4346
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5788376331329346,
+      "learning_rate": 4.8853555379328574e-05,
+      "loss": 1.3992,
+      "step": 4347
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9776932001113892,
+      "learning_rate": 4.8853291038858054e-05,
+      "loss": 1.5345,
+      "step": 4348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.273927927017212,
+      "learning_rate": 4.885302669838753e-05,
+      "loss": 1.5878,
+      "step": 4349
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4893364906311035,
+      "learning_rate": 4.8852762357917e-05,
+      "loss": 1.3743,
+      "step": 4350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6125288009643555,
+      "learning_rate": 4.8852498017446475e-05,
+      "loss": 1.3416,
+      "step": 4351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7043923139572144,
+      "learning_rate": 4.885223367697594e-05,
+      "loss": 1.3351,
+      "step": 4352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.104245185852051,
+      "learning_rate": 4.885196933650542e-05,
+      "loss": 1.8122,
+      "step": 4353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.839626431465149,
+      "learning_rate": 4.8851704996034895e-05,
+      "loss": 1.4431,
+      "step": 4354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.422130584716797,
+      "learning_rate": 4.885144065556437e-05,
+      "loss": 1.5603,
+      "step": 4355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.766993522644043,
+      "learning_rate": 4.885117631509384e-05,
+      "loss": 1.2945,
+      "step": 4356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5029704570770264,
+      "learning_rate": 4.8850911974623316e-05,
+      "loss": 1.3306,
+      "step": 4357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9337724447250366,
+      "learning_rate": 4.885064763415279e-05,
+      "loss": 1.4872,
+      "step": 4358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.32125186920166,
+      "learning_rate": 4.885038329368227e-05,
+      "loss": 1.6607,
+      "step": 4359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6352603435516357,
+      "learning_rate": 4.885011895321174e-05,
+      "loss": 1.3799,
+      "step": 4360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6401586532592773,
+      "learning_rate": 4.884985461274122e-05,
+      "loss": 1.509,
+      "step": 4361
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.013057231903076,
+      "learning_rate": 4.8849590272270684e-05,
+      "loss": 1.3349,
+      "step": 4362
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.963485598564148,
+      "learning_rate": 4.884932593180016e-05,
+      "loss": 1.4882,
+      "step": 4363
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3448379039764404,
+      "learning_rate": 4.884906159132964e-05,
+      "loss": 1.4742,
+      "step": 4364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.82082998752594,
+      "learning_rate": 4.8848797250859105e-05,
+      "loss": 1.5128,
+      "step": 4365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5631864070892334,
+      "learning_rate": 4.8848532910388585e-05,
+      "loss": 1.4164,
+      "step": 4366
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9738173484802246,
+      "learning_rate": 4.884826856991805e-05,
+      "loss": 1.5474,
+      "step": 4367
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8066543340682983,
+      "learning_rate": 4.884800422944753e-05,
+      "loss": 1.3015,
+      "step": 4368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.671018123626709,
+      "learning_rate": 4.8847739888977006e-05,
+      "loss": 1.5149,
+      "step": 4369
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6383161544799805,
+      "learning_rate": 4.884747554850648e-05,
+      "loss": 1.4153,
+      "step": 4370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6217504739761353,
+      "learning_rate": 4.884721120803595e-05,
+      "loss": 1.3047,
+      "step": 4371
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6231732368469238,
+      "learning_rate": 4.8846946867565426e-05,
+      "loss": 1.3857,
+      "step": 4372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7754180431365967,
+      "learning_rate": 4.88466825270949e-05,
+      "loss": 1.3169,
+      "step": 4373
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3648816347122192,
+      "learning_rate": 4.8846418186624374e-05,
+      "loss": 1.2077,
+      "step": 4374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8391146659851074,
+      "learning_rate": 4.884615384615385e-05,
+      "loss": 1.5896,
+      "step": 4375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.053676605224609,
+      "learning_rate": 4.884588950568332e-05,
+      "loss": 1.7556,
+      "step": 4376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.575204849243164,
+      "learning_rate": 4.8845625165212794e-05,
+      "loss": 1.362,
+      "step": 4377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.306635618209839,
+      "learning_rate": 4.884536082474227e-05,
+      "loss": 1.4635,
+      "step": 4378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.638535499572754,
+      "learning_rate": 4.884509648427175e-05,
+      "loss": 1.3317,
+      "step": 4379
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.59475839138031,
+      "learning_rate": 4.8844832143801215e-05,
+      "loss": 1.4167,
+      "step": 4380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4710384607315063,
+      "learning_rate": 4.8844567803330695e-05,
+      "loss": 1.3379,
+      "step": 4381
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6865142583847046,
+      "learning_rate": 4.884430346286016e-05,
+      "loss": 1.3806,
+      "step": 4382
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.861055612564087,
+      "learning_rate": 4.884403912238964e-05,
+      "loss": 1.3592,
+      "step": 4383
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.364955186843872,
+      "learning_rate": 4.8843774781919116e-05,
+      "loss": 1.5221,
+      "step": 4384
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.065742254257202,
+      "learning_rate": 4.884351044144859e-05,
+      "loss": 1.6041,
+      "step": 4385
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2970387935638428,
+      "learning_rate": 4.884324610097806e-05,
+      "loss": 1.6948,
+      "step": 4386
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6079778671264648,
+      "learning_rate": 4.884298176050754e-05,
+      "loss": 1.3136,
+      "step": 4387
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8889609575271606,
+      "learning_rate": 4.884271742003701e-05,
+      "loss": 1.4214,
+      "step": 4388
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.201957941055298,
+      "learning_rate": 4.8842453079566484e-05,
+      "loss": 1.4428,
+      "step": 4389
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9631575345993042,
+      "learning_rate": 4.884218873909596e-05,
+      "loss": 1.4521,
+      "step": 4390
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.002686023712158,
+      "learning_rate": 4.884192439862543e-05,
+      "loss": 1.3925,
+      "step": 4391
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8043475151062012,
+      "learning_rate": 4.8841660058154905e-05,
+      "loss": 1.3231,
+      "step": 4392
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5371848344802856,
+      "learning_rate": 4.884139571768438e-05,
+      "loss": 1.6754,
+      "step": 4393
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1253271102905273,
+      "learning_rate": 4.884113137721386e-05,
+      "loss": 1.43,
+      "step": 4394
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.800060749053955,
+      "learning_rate": 4.8840867036743325e-05,
+      "loss": 1.4813,
+      "step": 4395
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6781516075134277,
+      "learning_rate": 4.8840602696272806e-05,
+      "loss": 1.4111,
+      "step": 4396
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1759705543518066,
+      "learning_rate": 4.884033835580227e-05,
+      "loss": 1.558,
+      "step": 4397
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8059278726577759,
+      "learning_rate": 4.8840074015331746e-05,
+      "loss": 1.3979,
+      "step": 4398
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7733122110366821,
+      "learning_rate": 4.8839809674861226e-05,
+      "loss": 1.5473,
+      "step": 4399
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8667798042297363,
+      "learning_rate": 4.883954533439069e-05,
+      "loss": 1.4502,
+      "step": 4400
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3362619876861572,
+      "learning_rate": 4.8839280993920173e-05,
+      "loss": 2.0705,
+      "step": 4401
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.05210542678833,
+      "learning_rate": 4.883901665344964e-05,
+      "loss": 1.3842,
+      "step": 4402
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5056041479110718,
+      "learning_rate": 4.883875231297912e-05,
+      "loss": 1.3245,
+      "step": 4403
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2423758506774902,
+      "learning_rate": 4.8838487972508594e-05,
+      "loss": 1.3608,
+      "step": 4404
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0578110218048096,
+      "learning_rate": 4.883822363203807e-05,
+      "loss": 1.4167,
+      "step": 4405
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7352017164230347,
+      "learning_rate": 4.883795929156754e-05,
+      "loss": 1.4033,
+      "step": 4406
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1025049686431885,
+      "learning_rate": 4.8837694951097015e-05,
+      "loss": 1.4004,
+      "step": 4407
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.740944504737854,
+      "learning_rate": 4.883743061062649e-05,
+      "loss": 1.6138,
+      "step": 4408
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9742705821990967,
+      "learning_rate": 4.883716627015596e-05,
+      "loss": 1.5065,
+      "step": 4409
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.127041816711426,
+      "learning_rate": 4.8836901929685436e-05,
+      "loss": 1.4165,
+      "step": 4410
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4437282085418701,
+      "learning_rate": 4.883663758921491e-05,
+      "loss": 1.3417,
+      "step": 4411
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6292411088943481,
+      "learning_rate": 4.883637324874438e-05,
+      "loss": 1.5006,
+      "step": 4412
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.652422547340393,
+      "learning_rate": 4.8836108908273856e-05,
+      "loss": 1.2715,
+      "step": 4413
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4947385787963867,
+      "learning_rate": 4.883584456780334e-05,
+      "loss": 1.2692,
+      "step": 4414
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.61137056350708,
+      "learning_rate": 4.8835580227332803e-05,
+      "loss": 1.3911,
+      "step": 4415
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5426421165466309,
+      "learning_rate": 4.8835315886862284e-05,
+      "loss": 1.4165,
+      "step": 4416
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5341547727584839,
+      "learning_rate": 4.883505154639175e-05,
+      "loss": 1.3929,
+      "step": 4417
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5776658058166504,
+      "learning_rate": 4.883478720592123e-05,
+      "loss": 1.2885,
+      "step": 4418
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5758557319641113,
+      "learning_rate": 4.8834522865450705e-05,
+      "loss": 1.3348,
+      "step": 4419
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6359872817993164,
+      "learning_rate": 4.883425852498018e-05,
+      "loss": 1.372,
+      "step": 4420
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.516677975654602,
+      "learning_rate": 4.883399418450965e-05,
+      "loss": 1.3311,
+      "step": 4421
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.857268214225769,
+      "learning_rate": 4.8833729844039125e-05,
+      "loss": 1.9379,
+      "step": 4422
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6571722030639648,
+      "learning_rate": 4.88334655035686e-05,
+      "loss": 1.5929,
+      "step": 4423
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.181668519973755,
+      "learning_rate": 4.883320116309807e-05,
+      "loss": 1.488,
+      "step": 4424
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5646770000457764,
+      "learning_rate": 4.8832936822627546e-05,
+      "loss": 1.3466,
+      "step": 4425
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4397339820861816,
+      "learning_rate": 4.883267248215702e-05,
+      "loss": 1.5194,
+      "step": 4426
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4610471725463867,
+      "learning_rate": 4.883240814168649e-05,
+      "loss": 1.6237,
+      "step": 4427
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0211644172668457,
+      "learning_rate": 4.883214380121597e-05,
+      "loss": 1.3741,
+      "step": 4428
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5395281314849854,
+      "learning_rate": 4.883187946074545e-05,
+      "loss": 1.2766,
+      "step": 4429
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4669543504714966,
+      "learning_rate": 4.8831615120274914e-05,
+      "loss": 2.2489,
+      "step": 4430
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.70746910572052,
+      "learning_rate": 4.8831350779804394e-05,
+      "loss": 1.3987,
+      "step": 4431
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9694814682006836,
+      "learning_rate": 4.883108643933386e-05,
+      "loss": 1.479,
+      "step": 4432
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8405081033706665,
+      "learning_rate": 4.883082209886334e-05,
+      "loss": 1.3034,
+      "step": 4433
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6236685514450073,
+      "learning_rate": 4.8830557758392815e-05,
+      "loss": 1.283,
+      "step": 4434
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.070608615875244,
+      "learning_rate": 4.883029341792228e-05,
+      "loss": 1.3261,
+      "step": 4435
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6274160146713257,
+      "learning_rate": 4.883002907745176e-05,
+      "loss": 1.4248,
+      "step": 4436
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8484519720077515,
+      "learning_rate": 4.882976473698123e-05,
+      "loss": 1.4891,
+      "step": 4437
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5411738157272339,
+      "learning_rate": 4.882950039651071e-05,
+      "loss": 1.4084,
+      "step": 4438
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4379396438598633,
+      "learning_rate": 4.882923605604018e-05,
+      "loss": 1.5107,
+      "step": 4439
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.011383056640625,
+      "learning_rate": 4.8828971715569656e-05,
+      "loss": 1.3219,
+      "step": 4440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1692140102386475,
+      "learning_rate": 4.882870737509913e-05,
+      "loss": 1.5264,
+      "step": 4441
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7373038530349731,
+      "learning_rate": 4.8828443034628603e-05,
+      "loss": 1.5139,
+      "step": 4442
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0216622352600098,
+      "learning_rate": 4.882817869415808e-05,
+      "loss": 1.4549,
+      "step": 4443
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9556533098220825,
+      "learning_rate": 4.882791435368755e-05,
+      "loss": 1.4818,
+      "step": 4444
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7612721920013428,
+      "learning_rate": 4.8827650013217024e-05,
+      "loss": 1.3849,
+      "step": 4445
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7546637058258057,
+      "learning_rate": 4.88273856727465e-05,
+      "loss": 1.4027,
+      "step": 4446
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7115594148635864,
+      "learning_rate": 4.882712133227597e-05,
+      "loss": 1.3988,
+      "step": 4447
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8957990407943726,
+      "learning_rate": 4.8826856991805445e-05,
+      "loss": 1.4768,
+      "step": 4448
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.547739863395691,
+      "learning_rate": 4.8826592651334925e-05,
+      "loss": 1.5457,
+      "step": 4449
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.711211085319519,
+      "learning_rate": 4.882632831086439e-05,
+      "loss": 1.3561,
+      "step": 4450
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0038018226623535,
+      "learning_rate": 4.882606397039387e-05,
+      "loss": 1.3117,
+      "step": 4451
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7332072257995605,
+      "learning_rate": 4.882579962992334e-05,
+      "loss": 1.7041,
+      "step": 4452
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1292214393615723,
+      "learning_rate": 4.882553528945282e-05,
+      "loss": 1.5303,
+      "step": 4453
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0809857845306396,
+      "learning_rate": 4.882527094898229e-05,
+      "loss": 1.3886,
+      "step": 4454
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8059511184692383,
+      "learning_rate": 4.882500660851177e-05,
+      "loss": 1.3638,
+      "step": 4455
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.590413808822632,
+      "learning_rate": 4.882474226804124e-05,
+      "loss": 1.3949,
+      "step": 4456
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.916711688041687,
+      "learning_rate": 4.8824477927570714e-05,
+      "loss": 1.4233,
+      "step": 4457
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8072088956832886,
+      "learning_rate": 4.882421358710019e-05,
+      "loss": 1.4395,
+      "step": 4458
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5687143802642822,
+      "learning_rate": 4.882394924662966e-05,
+      "loss": 1.3741,
+      "step": 4459
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8665257692337036,
+      "learning_rate": 4.8823684906159134e-05,
+      "loss": 1.4136,
+      "step": 4460
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9107766151428223,
+      "learning_rate": 4.882342056568861e-05,
+      "loss": 1.5163,
+      "step": 4461
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5992307662963867,
+      "learning_rate": 4.882315622521808e-05,
+      "loss": 1.6445,
+      "step": 4462
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6568931341171265,
+      "learning_rate": 4.8822891884747555e-05,
+      "loss": 1.404,
+      "step": 4463
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.174093723297119,
+      "learning_rate": 4.8822627544277036e-05,
+      "loss": 1.5543,
+      "step": 4464
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.970145344734192,
+      "learning_rate": 4.88223632038065e-05,
+      "loss": 1.2664,
+      "step": 4465
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6569005250930786,
+      "learning_rate": 4.882209886333598e-05,
+      "loss": 1.3722,
+      "step": 4466
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0356826782226562,
+      "learning_rate": 4.882183452286545e-05,
+      "loss": 1.3992,
+      "step": 4467
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6675574779510498,
+      "learning_rate": 4.882157018239493e-05,
+      "loss": 1.4396,
+      "step": 4468
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6541173458099365,
+      "learning_rate": 4.8821305841924403e-05,
+      "loss": 1.4049,
+      "step": 4469
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7723355293273926,
+      "learning_rate": 4.882104150145387e-05,
+      "loss": 1.3982,
+      "step": 4470
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0055694580078125,
+      "learning_rate": 4.882077716098335e-05,
+      "loss": 1.5527,
+      "step": 4471
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9685750007629395,
+      "learning_rate": 4.882051282051282e-05,
+      "loss": 1.4427,
+      "step": 4472
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0867955684661865,
+      "learning_rate": 4.88202484800423e-05,
+      "loss": 1.574,
+      "step": 4473
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6283562183380127,
+      "learning_rate": 4.881998413957177e-05,
+      "loss": 1.9701,
+      "step": 4474
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.250705718994141,
+      "learning_rate": 4.8819719799101245e-05,
+      "loss": 1.4474,
+      "step": 4475
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.705241322517395,
+      "learning_rate": 4.881945545863072e-05,
+      "loss": 1.3599,
+      "step": 4476
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5406692028045654,
+      "learning_rate": 4.881919111816019e-05,
+      "loss": 1.3581,
+      "step": 4477
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0474462509155273,
+      "learning_rate": 4.8818926777689666e-05,
+      "loss": 1.5163,
+      "step": 4478
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5778311491012573,
+      "learning_rate": 4.8818662437219146e-05,
+      "loss": 1.3504,
+      "step": 4479
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4876577854156494,
+      "learning_rate": 4.881839809674861e-05,
+      "loss": 1.5687,
+      "step": 4480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.821390390396118,
+      "learning_rate": 4.8818133756278086e-05,
+      "loss": 1.4732,
+      "step": 4481
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6455339193344116,
+      "learning_rate": 4.881786941580756e-05,
+      "loss": 1.4398,
+      "step": 4482
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6818114519119263,
+      "learning_rate": 4.8817605075337033e-05,
+      "loss": 1.4049,
+      "step": 4483
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.683823823928833,
+      "learning_rate": 4.8817340734866514e-05,
+      "loss": 1.6756,
+      "step": 4484
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8038464784622192,
+      "learning_rate": 4.881707639439598e-05,
+      "loss": 1.4569,
+      "step": 4485
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8012572526931763,
+      "learning_rate": 4.881681205392546e-05,
+      "loss": 1.3464,
+      "step": 4486
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9295200109481812,
+      "learning_rate": 4.881654771345493e-05,
+      "loss": 2.0157,
+      "step": 4487
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.77413010597229,
+      "learning_rate": 4.881628337298441e-05,
+      "loss": 1.5822,
+      "step": 4488
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6211590766906738,
+      "learning_rate": 4.881601903251388e-05,
+      "loss": 1.3462,
+      "step": 4489
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7710767984390259,
+      "learning_rate": 4.8815754692043355e-05,
+      "loss": 1.4885,
+      "step": 4490
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9360524415969849,
+      "learning_rate": 4.881549035157283e-05,
+      "loss": 1.5318,
+      "step": 4491
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.461181640625,
+      "learning_rate": 4.88152260111023e-05,
+      "loss": 1.2711,
+      "step": 4492
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5264586210250854,
+      "learning_rate": 4.8814961670631776e-05,
+      "loss": 1.2443,
+      "step": 4493
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.596569299697876,
+      "learning_rate": 4.881469733016125e-05,
+      "loss": 1.4093,
+      "step": 4494
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5176851749420166,
+      "learning_rate": 4.881443298969072e-05,
+      "loss": 1.6282,
+      "step": 4495
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5386582612991333,
+      "learning_rate": 4.8814168649220197e-05,
+      "loss": 1.26,
+      "step": 4496
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6517564058303833,
+      "learning_rate": 4.881390430874967e-05,
+      "loss": 1.4419,
+      "step": 4497
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9463257789611816,
+      "learning_rate": 4.8813639968279144e-05,
+      "loss": 1.411,
+      "step": 4498
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3600172996520996,
+      "learning_rate": 4.8813375627808624e-05,
+      "loss": 1.6017,
+      "step": 4499
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0365679264068604,
+      "learning_rate": 4.881311128733809e-05,
+      "loss": 1.4694,
+      "step": 4500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.524747371673584,
+      "learning_rate": 4.881284694686757e-05,
+      "loss": 1.1768,
+      "step": 4501
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.016408920288086,
+      "learning_rate": 4.881258260639704e-05,
+      "loss": 1.4297,
+      "step": 4502
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4912422895431519,
+      "learning_rate": 4.881231826592652e-05,
+      "loss": 1.5179,
+      "step": 4503
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3293849229812622,
+      "learning_rate": 4.881205392545599e-05,
+      "loss": 1.2578,
+      "step": 4504
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.564589023590088,
+      "learning_rate": 4.8811789584985466e-05,
+      "loss": 1.4759,
+      "step": 4505
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3390494585037231,
+      "learning_rate": 4.881152524451494e-05,
+      "loss": 1.2973,
+      "step": 4506
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6192891597747803,
+      "learning_rate": 4.8811260904044406e-05,
+      "loss": 1.2618,
+      "step": 4507
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8427395820617676,
+      "learning_rate": 4.8810996563573886e-05,
+      "loss": 1.5455,
+      "step": 4508
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.765678882598877,
+      "learning_rate": 4.881073222310336e-05,
+      "loss": 1.6087,
+      "step": 4509
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7664467096328735,
+      "learning_rate": 4.881046788263283e-05,
+      "loss": 1.341,
+      "step": 4510
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9165674448013306,
+      "learning_rate": 4.881020354216231e-05,
+      "loss": 1.4437,
+      "step": 4511
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4680993556976318,
+      "learning_rate": 4.880993920169178e-05,
+      "loss": 1.439,
+      "step": 4512
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9422781467437744,
+      "learning_rate": 4.8809674861221254e-05,
+      "loss": 1.4193,
+      "step": 4513
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.20975661277771,
+      "learning_rate": 4.8809410520750734e-05,
+      "loss": 1.5039,
+      "step": 4514
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.543947458267212,
+      "learning_rate": 4.88091461802802e-05,
+      "loss": 1.4937,
+      "step": 4515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2626608610153198,
+      "learning_rate": 4.8808881839809675e-05,
+      "loss": 1.813,
+      "step": 4516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8492497205734253,
+      "learning_rate": 4.880861749933915e-05,
+      "loss": 1.4937,
+      "step": 4517
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5874241590499878,
+      "learning_rate": 4.880835315886862e-05,
+      "loss": 1.2873,
+      "step": 4518
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9864928722381592,
+      "learning_rate": 4.88080888183981e-05,
+      "loss": 1.4642,
+      "step": 4519
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9390459060668945,
+      "learning_rate": 4.880782447792757e-05,
+      "loss": 1.5054,
+      "step": 4520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4777390956878662,
+      "learning_rate": 4.880756013745705e-05,
+      "loss": 2.1588,
+      "step": 4521
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4482314586639404,
+      "learning_rate": 4.8807295796986516e-05,
+      "loss": 1.3556,
+      "step": 4522
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6229742765426636,
+      "learning_rate": 4.8807031456515997e-05,
+      "loss": 1.5875,
+      "step": 4523
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7337441444396973,
+      "learning_rate": 4.880676711604547e-05,
+      "loss": 1.3388,
+      "step": 4524
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9591526985168457,
+      "learning_rate": 4.8806502775574944e-05,
+      "loss": 1.5188,
+      "step": 4525
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8399215936660767,
+      "learning_rate": 4.880623843510442e-05,
+      "loss": 1.4325,
+      "step": 4526
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7989517450332642,
+      "learning_rate": 4.880597409463389e-05,
+      "loss": 1.3831,
+      "step": 4527
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6329283714294434,
+      "learning_rate": 4.8805709754163364e-05,
+      "loss": 1.4908,
+      "step": 4528
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5692601203918457,
+      "learning_rate": 4.880544541369284e-05,
+      "loss": 1.2804,
+      "step": 4529
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0226197242736816,
+      "learning_rate": 4.880518107322231e-05,
+      "loss": 1.363,
+      "step": 4530
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9411745071411133,
+      "learning_rate": 4.8804916732751785e-05,
+      "loss": 1.3867,
+      "step": 4531
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1090807914733887,
+      "learning_rate": 4.880465239228126e-05,
+      "loss": 1.5384,
+      "step": 4532
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8642851114273071,
+      "learning_rate": 4.880438805181073e-05,
+      "loss": 1.5675,
+      "step": 4533
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5542536973953247,
+      "learning_rate": 4.880412371134021e-05,
+      "loss": 1.2382,
+      "step": 4534
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5389326810836792,
+      "learning_rate": 4.880385937086968e-05,
+      "loss": 1.4613,
+      "step": 4535
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6144264936447144,
+      "learning_rate": 4.880359503039916e-05,
+      "loss": 1.4008,
+      "step": 4536
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9708188772201538,
+      "learning_rate": 4.8803330689928627e-05,
+      "loss": 1.4182,
+      "step": 4537
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.022989511489868,
+      "learning_rate": 4.880306634945811e-05,
+      "loss": 1.6196,
+      "step": 4538
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.044689178466797,
+      "learning_rate": 4.880280200898758e-05,
+      "loss": 1.4183,
+      "step": 4539
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7205331325531006,
+      "learning_rate": 4.8802537668517054e-05,
+      "loss": 1.3333,
+      "step": 4540
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.393510580062866,
+      "learning_rate": 4.880227332804653e-05,
+      "loss": 1.5581,
+      "step": 4541
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.54645836353302,
+      "learning_rate": 4.8802008987576e-05,
+      "loss": 1.2737,
+      "step": 4542
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6581310033798218,
+      "learning_rate": 4.8801744647105475e-05,
+      "loss": 1.3211,
+      "step": 4543
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.38960599899292,
+      "learning_rate": 4.880148030663495e-05,
+      "loss": 1.4274,
+      "step": 4544
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8591551780700684,
+      "learning_rate": 4.880121596616442e-05,
+      "loss": 1.5278,
+      "step": 4545
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5368527173995972,
+      "learning_rate": 4.8800951625693895e-05,
+      "loss": 1.4095,
+      "step": 4546
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5318002700805664,
+      "learning_rate": 4.880068728522337e-05,
+      "loss": 1.4377,
+      "step": 4547
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0156326293945312,
+      "learning_rate": 4.880042294475284e-05,
+      "loss": 1.5163,
+      "step": 4548
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6280224323272705,
+      "learning_rate": 4.880015860428232e-05,
+      "loss": 1.4189,
+      "step": 4549
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.485221028327942,
+      "learning_rate": 4.879989426381179e-05,
+      "loss": 1.308,
+      "step": 4550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4089244604110718,
+      "learning_rate": 4.879962992334127e-05,
+      "loss": 1.3689,
+      "step": 4551
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7515593767166138,
+      "learning_rate": 4.879936558287074e-05,
+      "loss": 1.3893,
+      "step": 4552
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7118414640426636,
+      "learning_rate": 4.879910124240021e-05,
+      "loss": 1.4796,
+      "step": 4553
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8423737287521362,
+      "learning_rate": 4.879883690192969e-05,
+      "loss": 1.3529,
+      "step": 4554
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5409581661224365,
+      "learning_rate": 4.879857256145916e-05,
+      "loss": 1.1308,
+      "step": 4555
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5676872730255127,
+      "learning_rate": 4.879830822098864e-05,
+      "loss": 1.3744,
+      "step": 4556
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.78885018825531,
+      "learning_rate": 4.8798043880518105e-05,
+      "loss": 1.3089,
+      "step": 4557
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9444341659545898,
+      "learning_rate": 4.8797779540047585e-05,
+      "loss": 1.3942,
+      "step": 4558
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4420915842056274,
+      "learning_rate": 4.879751519957706e-05,
+      "loss": 1.2625,
+      "step": 4559
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8153698444366455,
+      "learning_rate": 4.879725085910653e-05,
+      "loss": 1.2902,
+      "step": 4560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0951809883117676,
+      "learning_rate": 4.8796986518636006e-05,
+      "loss": 1.434,
+      "step": 4561
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1119327545166016,
+      "learning_rate": 4.879672217816548e-05,
+      "loss": 1.5159,
+      "step": 4562
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8318088054656982,
+      "learning_rate": 4.879645783769495e-05,
+      "loss": 1.6037,
+      "step": 4563
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.564026117324829,
+      "learning_rate": 4.8796193497224427e-05,
+      "loss": 1.4926,
+      "step": 4564
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9186251163482666,
+      "learning_rate": 4.87959291567539e-05,
+      "loss": 1.5834,
+      "step": 4565
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5816538333892822,
+      "learning_rate": 4.8795664816283374e-05,
+      "loss": 1.3721,
+      "step": 4566
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.469501256942749,
+      "learning_rate": 4.879540047581285e-05,
+      "loss": 1.3687,
+      "step": 4567
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9424313306808472,
+      "learning_rate": 4.879513613534232e-05,
+      "loss": 1.5072,
+      "step": 4568
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3936269283294678,
+      "learning_rate": 4.87948717948718e-05,
+      "loss": 1.5672,
+      "step": 4569
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1221189498901367,
+      "learning_rate": 4.879460745440127e-05,
+      "loss": 1.3303,
+      "step": 4570
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7698054313659668,
+      "learning_rate": 4.879434311393075e-05,
+      "loss": 1.4831,
+      "step": 4571
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7982372045516968,
+      "learning_rate": 4.8794078773460215e-05,
+      "loss": 1.3622,
+      "step": 4572
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2534544467926025,
+      "learning_rate": 4.8793814432989695e-05,
+      "loss": 1.3632,
+      "step": 4573
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6845695972442627,
+      "learning_rate": 4.879355009251917e-05,
+      "loss": 1.4178,
+      "step": 4574
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.139815330505371,
+      "learning_rate": 4.879328575204864e-05,
+      "loss": 1.496,
+      "step": 4575
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8460251092910767,
+      "learning_rate": 4.8793021411578116e-05,
+      "loss": 1.3391,
+      "step": 4576
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.672227144241333,
+      "learning_rate": 4.879275707110759e-05,
+      "loss": 1.4704,
+      "step": 4577
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.633679747581482,
+      "learning_rate": 4.879249273063706e-05,
+      "loss": 1.3619,
+      "step": 4578
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4078054428100586,
+      "learning_rate": 4.879222839016654e-05,
+      "loss": 1.4268,
+      "step": 4579
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4188555479049683,
+      "learning_rate": 4.879196404969601e-05,
+      "loss": 1.7585,
+      "step": 4580
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5723299980163574,
+      "learning_rate": 4.8791699709225484e-05,
+      "loss": 1.335,
+      "step": 4581
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9830964803695679,
+      "learning_rate": 4.879143536875496e-05,
+      "loss": 1.3924,
+      "step": 4582
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.125983953475952,
+      "learning_rate": 4.879117102828443e-05,
+      "loss": 1.3872,
+      "step": 4583
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.534693956375122,
+      "learning_rate": 4.879090668781391e-05,
+      "loss": 1.4463,
+      "step": 4584
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.48050057888031,
+      "learning_rate": 4.879064234734338e-05,
+      "loss": 1.2942,
+      "step": 4585
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0022566318511963,
+      "learning_rate": 4.879037800687286e-05,
+      "loss": 1.5049,
+      "step": 4586
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.632033109664917,
+      "learning_rate": 4.8790113666402325e-05,
+      "loss": 2.1048,
+      "step": 4587
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6605455875396729,
+      "learning_rate": 4.87898493259318e-05,
+      "loss": 1.5114,
+      "step": 4588
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6955710649490356,
+      "learning_rate": 4.878958498546128e-05,
+      "loss": 1.4387,
+      "step": 4589
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6719352006912231,
+      "learning_rate": 4.8789320644990746e-05,
+      "loss": 1.8874,
+      "step": 4590
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.799495816230774,
+      "learning_rate": 4.8789056304520226e-05,
+      "loss": 1.4929,
+      "step": 4591
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6134940385818481,
+      "learning_rate": 4.878879196404969e-05,
+      "loss": 1.3948,
+      "step": 4592
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4695196151733398,
+      "learning_rate": 4.8788527623579174e-05,
+      "loss": 1.3337,
+      "step": 4593
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.06177020072937,
+      "learning_rate": 4.878826328310865e-05,
+      "loss": 1.5201,
+      "step": 4594
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7044192552566528,
+      "learning_rate": 4.878799894263812e-05,
+      "loss": 1.4849,
+      "step": 4595
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5053993463516235,
+      "learning_rate": 4.8787734602167594e-05,
+      "loss": 1.3265,
+      "step": 4596
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8545057773590088,
+      "learning_rate": 4.878747026169707e-05,
+      "loss": 1.5414,
+      "step": 4597
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8563995361328125,
+      "learning_rate": 4.878720592122654e-05,
+      "loss": 1.4112,
+      "step": 4598
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7495529651641846,
+      "learning_rate": 4.8786941580756015e-05,
+      "loss": 1.502,
+      "step": 4599
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5158578157424927,
+      "learning_rate": 4.878667724028549e-05,
+      "loss": 1.3194,
+      "step": 4600
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.579588770866394,
+      "learning_rate": 4.878641289981496e-05,
+      "loss": 1.3677,
+      "step": 4601
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6226986646652222,
+      "learning_rate": 4.8786148559344436e-05,
+      "loss": 1.3915,
+      "step": 4602
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4236507415771484,
+      "learning_rate": 4.878588421887391e-05,
+      "loss": 1.4246,
+      "step": 4603
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5019302368164062,
+      "learning_rate": 4.878561987840339e-05,
+      "loss": 1.5155,
+      "step": 4604
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4651440382003784,
+      "learning_rate": 4.8785355537932856e-05,
+      "loss": 1.3555,
+      "step": 4605
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.355532169342041,
+      "learning_rate": 4.878509119746234e-05,
+      "loss": 1.3677,
+      "step": 4606
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7746658325195312,
+      "learning_rate": 4.8784826856991804e-05,
+      "loss": 1.5309,
+      "step": 4607
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6465436220169067,
+      "learning_rate": 4.8784562516521284e-05,
+      "loss": 1.4485,
+      "step": 4608
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4998319149017334,
+      "learning_rate": 4.878429817605076e-05,
+      "loss": 2.2349,
+      "step": 4609
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8081246614456177,
+      "learning_rate": 4.878403383558023e-05,
+      "loss": 1.6029,
+      "step": 4610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5944006443023682,
+      "learning_rate": 4.8783769495109705e-05,
+      "loss": 1.516,
+      "step": 4611
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6547672748565674,
+      "learning_rate": 4.878350515463918e-05,
+      "loss": 1.4337,
+      "step": 4612
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5510098934173584,
+      "learning_rate": 4.878324081416865e-05,
+      "loss": 1.3144,
+      "step": 4613
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.704793930053711,
+      "learning_rate": 4.8782976473698125e-05,
+      "loss": 1.1996,
+      "step": 4614
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.748790979385376,
+      "learning_rate": 4.87827121332276e-05,
+      "loss": 1.356,
+      "step": 4615
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5467334985733032,
+      "learning_rate": 4.878244779275707e-05,
+      "loss": 1.3549,
+      "step": 4616
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8900293111801147,
+      "learning_rate": 4.8782183452286546e-05,
+      "loss": 1.4671,
+      "step": 4617
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6885595321655273,
+      "learning_rate": 4.878191911181602e-05,
+      "loss": 1.5709,
+      "step": 4618
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.764149785041809,
+      "learning_rate": 4.87816547713455e-05,
+      "loss": 1.3872,
+      "step": 4619
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2067177295684814,
+      "learning_rate": 4.878139043087497e-05,
+      "loss": 1.4618,
+      "step": 4620
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.880245327949524,
+      "learning_rate": 4.878112609040445e-05,
+      "loss": 1.3161,
+      "step": 4621
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.054014205932617,
+      "learning_rate": 4.8780861749933914e-05,
+      "loss": 1.4202,
+      "step": 4622
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2725980281829834,
+      "learning_rate": 4.8780597409463394e-05,
+      "loss": 1.7349,
+      "step": 4623
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.195603132247925,
+      "learning_rate": 4.878033306899287e-05,
+      "loss": 1.4426,
+      "step": 4624
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.004899263381958,
+      "learning_rate": 4.8780068728522335e-05,
+      "loss": 1.3145,
+      "step": 4625
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8065818548202515,
+      "learning_rate": 4.8779804388051815e-05,
+      "loss": 1.5619,
+      "step": 4626
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6242791414260864,
+      "learning_rate": 4.877954004758128e-05,
+      "loss": 2.208,
+      "step": 4627
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5256190299987793,
+      "learning_rate": 4.877927570711076e-05,
+      "loss": 1.3063,
+      "step": 4628
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5420897006988525,
+      "learning_rate": 4.8779011366640236e-05,
+      "loss": 1.2783,
+      "step": 4629
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5026754140853882,
+      "learning_rate": 4.877874702616971e-05,
+      "loss": 1.3498,
+      "step": 4630
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8266685009002686,
+      "learning_rate": 4.877848268569918e-05,
+      "loss": 1.4188,
+      "step": 4631
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.943894863128662,
+      "learning_rate": 4.8778218345228656e-05,
+      "loss": 1.4898,
+      "step": 4632
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.63317608833313,
+      "learning_rate": 4.877795400475813e-05,
+      "loss": 1.5238,
+      "step": 4633
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.82279372215271,
+      "learning_rate": 4.8777689664287604e-05,
+      "loss": 1.394,
+      "step": 4634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.421159029006958,
+      "learning_rate": 4.877742532381708e-05,
+      "loss": 1.6817,
+      "step": 4635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.565055012702942,
+      "learning_rate": 4.877716098334655e-05,
+      "loss": 1.3231,
+      "step": 4636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7319974899291992,
+      "learning_rate": 4.8776896642876024e-05,
+      "loss": 1.6049,
+      "step": 4637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.643203616142273,
+      "learning_rate": 4.87766323024055e-05,
+      "loss": 1.2701,
+      "step": 4638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0534443855285645,
+      "learning_rate": 4.877636796193498e-05,
+      "loss": 1.4415,
+      "step": 4639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8320326805114746,
+      "learning_rate": 4.8776103621464445e-05,
+      "loss": 1.4298,
+      "step": 4640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2428292036056519,
+      "learning_rate": 4.8775839280993925e-05,
+      "loss": 1.2293,
+      "step": 4641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.690855026245117,
+      "learning_rate": 4.877557494052339e-05,
+      "loss": 1.4959,
+      "step": 4642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7312800884246826,
+      "learning_rate": 4.877531060005287e-05,
+      "loss": 1.4273,
+      "step": 4643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.966371536254883,
+      "learning_rate": 4.8775046259582346e-05,
+      "loss": 1.3957,
+      "step": 4644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1522960662841797,
+      "learning_rate": 4.877478191911182e-05,
+      "loss": 1.5181,
+      "step": 4645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.888797640800476,
+      "learning_rate": 4.877451757864129e-05,
+      "loss": 1.4025,
+      "step": 4646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7078485488891602,
+      "learning_rate": 4.877425323817077e-05,
+      "loss": 1.3607,
+      "step": 4647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5692074298858643,
+      "learning_rate": 4.877398889770024e-05,
+      "loss": 1.4507,
+      "step": 4648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6192700862884521,
+      "learning_rate": 4.8773724557229714e-05,
+      "loss": 1.3823,
+      "step": 4649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6030677556991577,
+      "learning_rate": 4.877346021675919e-05,
+      "loss": 1.3787,
+      "step": 4650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.812659740447998,
+      "learning_rate": 4.877319587628866e-05,
+      "loss": 1.3692,
+      "step": 4651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5404444932937622,
+      "learning_rate": 4.8772931535818135e-05,
+      "loss": 1.3494,
+      "step": 4652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7926093339920044,
+      "learning_rate": 4.877266719534761e-05,
+      "loss": 1.4986,
+      "step": 4653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9020531177520752,
+      "learning_rate": 4.877240285487709e-05,
+      "loss": 1.5307,
+      "step": 4654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4169816970825195,
+      "learning_rate": 4.8772138514406555e-05,
+      "loss": 1.2484,
+      "step": 4655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7983547449111938,
+      "learning_rate": 4.8771874173936036e-05,
+      "loss": 1.5385,
+      "step": 4656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2258872985839844,
+      "learning_rate": 4.87716098334655e-05,
+      "loss": 1.5281,
+      "step": 4657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.132728099822998,
+      "learning_rate": 4.877134549299498e-05,
+      "loss": 1.4223,
+      "step": 4658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5889049768447876,
+      "learning_rate": 4.8771081152524456e-05,
+      "loss": 1.2649,
+      "step": 4659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5479423999786377,
+      "learning_rate": 4.877081681205393e-05,
+      "loss": 1.4195,
+      "step": 4660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9602155685424805,
+      "learning_rate": 4.8770552471583404e-05,
+      "loss": 1.4297,
+      "step": 4661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7044546604156494,
+      "learning_rate": 4.877028813111287e-05,
+      "loss": 1.3513,
+      "step": 4662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3372812271118164,
+      "learning_rate": 4.877002379064235e-05,
+      "loss": 1.411,
+      "step": 4663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.65136981010437,
+      "learning_rate": 4.8769759450171824e-05,
+      "loss": 1.4067,
+      "step": 4664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9458646774291992,
+      "learning_rate": 4.87694951097013e-05,
+      "loss": 1.4973,
+      "step": 4665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9947926998138428,
+      "learning_rate": 4.876923076923077e-05,
+      "loss": 1.8429,
+      "step": 4666
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.503195881843567,
+      "learning_rate": 4.8768966428760245e-05,
+      "loss": 1.4375,
+      "step": 4667
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.567908525466919,
+      "learning_rate": 4.876870208828972e-05,
+      "loss": 1.3796,
+      "step": 4668
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5349079370498657,
+      "learning_rate": 4.87684377478192e-05,
+      "loss": 1.4292,
+      "step": 4669
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6886627674102783,
+      "learning_rate": 4.8768173407348666e-05,
+      "loss": 1.3932,
+      "step": 4670
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.203519105911255,
+      "learning_rate": 4.876790906687814e-05,
+      "loss": 1.3917,
+      "step": 4671
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6871525049209595,
+      "learning_rate": 4.876764472640761e-05,
+      "loss": 2.1095,
+      "step": 4672
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9410178661346436,
+      "learning_rate": 4.8767380385937086e-05,
+      "loss": 1.2765,
+      "step": 4673
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7879878282546997,
+      "learning_rate": 4.876711604546657e-05,
+      "loss": 1.521,
+      "step": 4674
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8162987232208252,
+      "learning_rate": 4.8766851704996034e-05,
+      "loss": 1.4222,
+      "step": 4675
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.408370018005371,
+      "learning_rate": 4.8766587364525514e-05,
+      "loss": 2.0306,
+      "step": 4676
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7677292823791504,
+      "learning_rate": 4.876632302405498e-05,
+      "loss": 1.3254,
+      "step": 4677
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4280229806900024,
+      "learning_rate": 4.876605868358446e-05,
+      "loss": 1.3338,
+      "step": 4678
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6176737546920776,
+      "learning_rate": 4.8765794343113935e-05,
+      "loss": 1.4205,
+      "step": 4679
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5336419343948364,
+      "learning_rate": 4.876553000264341e-05,
+      "loss": 1.5169,
+      "step": 4680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1140832901000977,
+      "learning_rate": 4.876526566217288e-05,
+      "loss": 1.4261,
+      "step": 4681
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.388932466506958,
+      "learning_rate": 4.8765001321702355e-05,
+      "loss": 1.4875,
+      "step": 4682
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6996467113494873,
+      "learning_rate": 4.876473698123183e-05,
+      "loss": 1.4497,
+      "step": 4683
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1638360023498535,
+      "learning_rate": 4.87644726407613e-05,
+      "loss": 1.5098,
+      "step": 4684
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.497776508331299,
+      "learning_rate": 4.8764208300290776e-05,
+      "loss": 1.3051,
+      "step": 4685
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.531911015510559,
+      "learning_rate": 4.876394395982025e-05,
+      "loss": 1.4743,
+      "step": 4686
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8110852241516113,
+      "learning_rate": 4.876367961934972e-05,
+      "loss": 1.4759,
+      "step": 4687
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8512872457504272,
+      "learning_rate": 4.87634152788792e-05,
+      "loss": 1.4222,
+      "step": 4688
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1822049617767334,
+      "learning_rate": 4.876315093840868e-05,
+      "loss": 1.622,
+      "step": 4689
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9273500442504883,
+      "learning_rate": 4.8762886597938144e-05,
+      "loss": 1.5329,
+      "step": 4690
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1991846561431885,
+      "learning_rate": 4.8762622257467624e-05,
+      "loss": 1.4344,
+      "step": 4691
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5737574100494385,
+      "learning_rate": 4.876235791699709e-05,
+      "loss": 1.3406,
+      "step": 4692
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6933783292770386,
+      "learning_rate": 4.876209357652657e-05,
+      "loss": 1.2708,
+      "step": 4693
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.660699486732483,
+      "learning_rate": 4.8761829236056045e-05,
+      "loss": 2.4238,
+      "step": 4694
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.171765089035034,
+      "learning_rate": 4.876156489558552e-05,
+      "loss": 1.609,
+      "step": 4695
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5206811428070068,
+      "learning_rate": 4.876130055511499e-05,
+      "loss": 1.3442,
+      "step": 4696
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4160341024398804,
+      "learning_rate": 4.876103621464446e-05,
+      "loss": 1.3411,
+      "step": 4697
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.175623655319214,
+      "learning_rate": 4.876077187417394e-05,
+      "loss": 1.3222,
+      "step": 4698
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8552768230438232,
+      "learning_rate": 4.876050753370341e-05,
+      "loss": 1.5434,
+      "step": 4699
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5857428312301636,
+      "learning_rate": 4.8760243193232886e-05,
+      "loss": 1.4594,
+      "step": 4700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4911208152770996,
+      "learning_rate": 4.875997885276236e-05,
+      "loss": 1.3775,
+      "step": 4701
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.87409508228302,
+      "learning_rate": 4.8759714512291834e-05,
+      "loss": 1.5288,
+      "step": 4702
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7333972454071045,
+      "learning_rate": 4.875945017182131e-05,
+      "loss": 1.4116,
+      "step": 4703
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.39243745803833,
+      "learning_rate": 4.875918583135079e-05,
+      "loss": 1.4486,
+      "step": 4704
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4174115657806396,
+      "learning_rate": 4.8758921490880254e-05,
+      "loss": 1.6693,
+      "step": 4705
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.543480157852173,
+      "learning_rate": 4.875865715040973e-05,
+      "loss": 1.3988,
+      "step": 4706
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6617071628570557,
+      "learning_rate": 4.87583928099392e-05,
+      "loss": 1.3551,
+      "step": 4707
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3982293605804443,
+      "learning_rate": 4.8758128469468675e-05,
+      "loss": 1.6767,
+      "step": 4708
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4041476249694824,
+      "learning_rate": 4.8757864128998155e-05,
+      "loss": 1.6411,
+      "step": 4709
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6969548463821411,
+      "learning_rate": 4.875759978852762e-05,
+      "loss": 1.4278,
+      "step": 4710
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7072598934173584,
+      "learning_rate": 4.87573354480571e-05,
+      "loss": 1.4097,
+      "step": 4711
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3222849369049072,
+      "learning_rate": 4.875707110758657e-05,
+      "loss": 1.2965,
+      "step": 4712
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6285868883132935,
+      "learning_rate": 4.875680676711605e-05,
+      "loss": 1.3033,
+      "step": 4713
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.763828158378601,
+      "learning_rate": 4.875654242664552e-05,
+      "loss": 1.5354,
+      "step": 4714
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8290836811065674,
+      "learning_rate": 4.8756278086175e-05,
+      "loss": 1.5165,
+      "step": 4715
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.569624900817871,
+      "learning_rate": 4.875601374570447e-05,
+      "loss": 1.5223,
+      "step": 4716
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6751562356948853,
+      "learning_rate": 4.8755749405233944e-05,
+      "loss": 1.5038,
+      "step": 4717
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6569452285766602,
+      "learning_rate": 4.875548506476342e-05,
+      "loss": 1.4082,
+      "step": 4718
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.734844446182251,
+      "learning_rate": 4.875522072429289e-05,
+      "loss": 1.4994,
+      "step": 4719
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6988458633422852,
+      "learning_rate": 4.8754956383822365e-05,
+      "loss": 1.4507,
+      "step": 4720
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8062493801116943,
+      "learning_rate": 4.875469204335184e-05,
+      "loss": 1.3653,
+      "step": 4721
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6240191459655762,
+      "learning_rate": 4.875442770288131e-05,
+      "loss": 1.3572,
+      "step": 4722
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.69684636592865,
+      "learning_rate": 4.8754163362410785e-05,
+      "loss": 1.365,
+      "step": 4723
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6635380983352661,
+      "learning_rate": 4.8753899021940266e-05,
+      "loss": 1.4655,
+      "step": 4724
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.678979754447937,
+      "learning_rate": 4.875363468146973e-05,
+      "loss": 1.3465,
+      "step": 4725
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5357449054718018,
+      "learning_rate": 4.875337034099921e-05,
+      "loss": 1.4531,
+      "step": 4726
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6130661964416504,
+      "learning_rate": 4.875310600052868e-05,
+      "loss": 1.3151,
+      "step": 4727
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6574233770370483,
+      "learning_rate": 4.875284166005816e-05,
+      "loss": 1.1939,
+      "step": 4728
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5522215366363525,
+      "learning_rate": 4.8752577319587633e-05,
+      "loss": 1.467,
+      "step": 4729
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6513762474060059,
+      "learning_rate": 4.875231297911711e-05,
+      "loss": 1.4554,
+      "step": 4730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2946112155914307,
+      "learning_rate": 4.875204863864658e-05,
+      "loss": 1.4577,
+      "step": 4731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1119039058685303,
+      "learning_rate": 4.8751784298176054e-05,
+      "loss": 1.6656,
+      "step": 4732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7246100902557373,
+      "learning_rate": 4.875151995770553e-05,
+      "loss": 1.4624,
+      "step": 4733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6317287683486938,
+      "learning_rate": 4.8751255617235e-05,
+      "loss": 1.3589,
+      "step": 4734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1683403253555298,
+      "learning_rate": 4.8750991276764475e-05,
+      "loss": 1.2245,
+      "step": 4735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7752867937088013,
+      "learning_rate": 4.875072693629395e-05,
+      "loss": 1.4884,
+      "step": 4736
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6410491466522217,
+      "learning_rate": 4.875046259582342e-05,
+      "loss": 1.2844,
+      "step": 4737
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.30256986618042,
+      "learning_rate": 4.8750198255352896e-05,
+      "loss": 1.2357,
+      "step": 4738
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4332902431488037,
+      "learning_rate": 4.8749933914882376e-05,
+      "loss": 1.4744,
+      "step": 4739
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.429681658744812,
+      "learning_rate": 4.874966957441184e-05,
+      "loss": 1.5768,
+      "step": 4740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9472750425338745,
+      "learning_rate": 4.874940523394132e-05,
+      "loss": 1.5824,
+      "step": 4741
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1008732318878174,
+      "learning_rate": 4.874914089347079e-05,
+      "loss": 1.4934,
+      "step": 4742
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5448297262191772,
+      "learning_rate": 4.8748876553000263e-05,
+      "loss": 1.3124,
+      "step": 4743
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.398728370666504,
+      "learning_rate": 4.8748612212529744e-05,
+      "loss": 1.4659,
+      "step": 4744
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3195762634277344,
+      "learning_rate": 4.874834787205921e-05,
+      "loss": 1.3,
+      "step": 4745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0743842124938965,
+      "learning_rate": 4.874808353158869e-05,
+      "loss": 1.5487,
+      "step": 4746
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8819688558578491,
+      "learning_rate": 4.874781919111816e-05,
+      "loss": 1.5032,
+      "step": 4747
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.462782621383667,
+      "learning_rate": 4.874755485064764e-05,
+      "loss": 1.4149,
+      "step": 4748
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7018969058990479,
+      "learning_rate": 4.874729051017711e-05,
+      "loss": 2.1972,
+      "step": 4749
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5215851068496704,
+      "learning_rate": 4.8747026169706585e-05,
+      "loss": 1.4515,
+      "step": 4750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7581164836883545,
+      "learning_rate": 4.874676182923606e-05,
+      "loss": 1.4497,
+      "step": 4751
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6670103073120117,
+      "learning_rate": 4.874649748876553e-05,
+      "loss": 1.3708,
+      "step": 4752
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8532363176345825,
+      "learning_rate": 4.8746233148295006e-05,
+      "loss": 1.3925,
+      "step": 4753
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8353195190429688,
+      "learning_rate": 4.874596880782448e-05,
+      "loss": 1.7824,
+      "step": 4754
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5380983352661133,
+      "learning_rate": 4.874570446735395e-05,
+      "loss": 1.4006,
+      "step": 4755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.765323281288147,
+      "learning_rate": 4.874544012688343e-05,
+      "loss": 1.2532,
+      "step": 4756
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8918789625167847,
+      "learning_rate": 4.87451757864129e-05,
+      "loss": 1.3686,
+      "step": 4757
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6504881381988525,
+      "learning_rate": 4.8744911445942374e-05,
+      "loss": 1.4611,
+      "step": 4758
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8274656534194946,
+      "learning_rate": 4.8744647105471854e-05,
+      "loss": 1.5501,
+      "step": 4759
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.700167179107666,
+      "learning_rate": 4.874438276500132e-05,
+      "loss": 1.2786,
+      "step": 4760
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.017512083053589,
+      "learning_rate": 4.87441184245308e-05,
+      "loss": 1.3305,
+      "step": 4761
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.473363995552063,
+      "learning_rate": 4.874385408406027e-05,
+      "loss": 1.3693,
+      "step": 4762
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4652252197265625,
+      "learning_rate": 4.874358974358975e-05,
+      "loss": 1.4468,
+      "step": 4763
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.112929344177246,
+      "learning_rate": 4.874332540311922e-05,
+      "loss": 1.6363,
+      "step": 4764
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.150355100631714,
+      "learning_rate": 4.8743061062648696e-05,
+      "loss": 1.3474,
+      "step": 4765
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.523508906364441,
+      "learning_rate": 4.874279672217817e-05,
+      "loss": 1.428,
+      "step": 4766
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3372673988342285,
+      "learning_rate": 4.874253238170764e-05,
+      "loss": 1.3271,
+      "step": 4767
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8846659660339355,
+      "learning_rate": 4.8742268041237116e-05,
+      "loss": 1.4821,
+      "step": 4768
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7223566770553589,
+      "learning_rate": 4.874200370076659e-05,
+      "loss": 1.3911,
+      "step": 4769
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5898950099945068,
+      "learning_rate": 4.8741739360296063e-05,
+      "loss": 1.3848,
+      "step": 4770
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.455134630203247,
+      "learning_rate": 4.874147501982554e-05,
+      "loss": 1.495,
+      "step": 4771
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2490167617797852,
+      "learning_rate": 4.874121067935501e-05,
+      "loss": 1.5674,
+      "step": 4772
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6180646419525146,
+      "learning_rate": 4.8740946338884484e-05,
+      "loss": 1.2164,
+      "step": 4773
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8182560205459595,
+      "learning_rate": 4.8740681998413965e-05,
+      "loss": 1.3955,
+      "step": 4774
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7403271198272705,
+      "learning_rate": 4.874041765794343e-05,
+      "loss": 1.6568,
+      "step": 4775
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9532129764556885,
+      "learning_rate": 4.874015331747291e-05,
+      "loss": 1.2436,
+      "step": 4776
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5287152528762817,
+      "learning_rate": 4.873988897700238e-05,
+      "loss": 1.2497,
+      "step": 4777
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4654107093811035,
+      "learning_rate": 4.873962463653185e-05,
+      "loss": 1.4408,
+      "step": 4778
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5461516380310059,
+      "learning_rate": 4.873936029606133e-05,
+      "loss": 1.4737,
+      "step": 4779
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8584709167480469,
+      "learning_rate": 4.87390959555908e-05,
+      "loss": 1.3586,
+      "step": 4780
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7089766263961792,
+      "learning_rate": 4.873883161512028e-05,
+      "loss": 1.4346,
+      "step": 4781
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.755517601966858,
+      "learning_rate": 4.8738567274649746e-05,
+      "loss": 1.5216,
+      "step": 4782
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7623766660690308,
+      "learning_rate": 4.873830293417923e-05,
+      "loss": 1.4949,
+      "step": 4783
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4704782962799072,
+      "learning_rate": 4.87380385937087e-05,
+      "loss": 1.4878,
+      "step": 4784
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1682727336883545,
+      "learning_rate": 4.8737774253238174e-05,
+      "loss": 1.4294,
+      "step": 4785
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.049515962600708,
+      "learning_rate": 4.873750991276765e-05,
+      "loss": 1.4267,
+      "step": 4786
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2890818119049072,
+      "learning_rate": 4.873724557229712e-05,
+      "loss": 1.4189,
+      "step": 4787
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3046393394470215,
+      "learning_rate": 4.8736981231826595e-05,
+      "loss": 1.447,
+      "step": 4788
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7137200832366943,
+      "learning_rate": 4.873671689135607e-05,
+      "loss": 1.2508,
+      "step": 4789
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7224429845809937,
+      "learning_rate": 4.873645255088554e-05,
+      "loss": 1.2838,
+      "step": 4790
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6841915845870972,
+      "learning_rate": 4.8736188210415015e-05,
+      "loss": 1.3479,
+      "step": 4791
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.451516032218933,
+      "learning_rate": 4.873592386994449e-05,
+      "loss": 1.2865,
+      "step": 4792
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6055448055267334,
+      "learning_rate": 4.873565952947396e-05,
+      "loss": 1.5004,
+      "step": 4793
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.307981014251709,
+      "learning_rate": 4.873539518900344e-05,
+      "loss": 1.7045,
+      "step": 4794
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3108603954315186,
+      "learning_rate": 4.873513084853291e-05,
+      "loss": 1.4873,
+      "step": 4795
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6384363174438477,
+      "learning_rate": 4.873486650806239e-05,
+      "loss": 1.4663,
+      "step": 4796
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8172800540924072,
+      "learning_rate": 4.873460216759186e-05,
+      "loss": 1.3506,
+      "step": 4797
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6121656894683838,
+      "learning_rate": 4.873433782712134e-05,
+      "loss": 1.3224,
+      "step": 4798
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5530335903167725,
+      "learning_rate": 4.873407348665081e-05,
+      "loss": 1.3168,
+      "step": 4799
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9251124858856201,
+      "learning_rate": 4.8733809146180284e-05,
+      "loss": 1.3867,
+      "step": 4800
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4822475910186768,
+      "learning_rate": 4.873354480570976e-05,
+      "loss": 1.5314,
+      "step": 4801
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3821372985839844,
+      "learning_rate": 4.873328046523923e-05,
+      "loss": 1.2861,
+      "step": 4802
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4548232555389404,
+      "learning_rate": 4.8733016124768705e-05,
+      "loss": 1.3341,
+      "step": 4803
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.630623698234558,
+      "learning_rate": 4.873275178429818e-05,
+      "loss": 1.3056,
+      "step": 4804
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9671969413757324,
+      "learning_rate": 4.873248744382765e-05,
+      "loss": 1.392,
+      "step": 4805
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.810545563697815,
+      "learning_rate": 4.8732223103357126e-05,
+      "loss": 1.3078,
+      "step": 4806
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5690481662750244,
+      "learning_rate": 4.87319587628866e-05,
+      "loss": 1.6786,
+      "step": 4807
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7635518312454224,
+      "learning_rate": 4.873169442241607e-05,
+      "loss": 1.3744,
+      "step": 4808
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7273030281066895,
+      "learning_rate": 4.873143008194555e-05,
+      "loss": 1.3589,
+      "step": 4809
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.334484815597534,
+      "learning_rate": 4.873116574147502e-05,
+      "loss": 1.4639,
+      "step": 4810
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5034682750701904,
+      "learning_rate": 4.87309014010045e-05,
+      "loss": 1.5499,
+      "step": 4811
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8367975950241089,
+      "learning_rate": 4.873063706053397e-05,
+      "loss": 1.4709,
+      "step": 4812
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9784109592437744,
+      "learning_rate": 4.873037272006345e-05,
+      "loss": 1.5397,
+      "step": 4813
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.662536382675171,
+      "learning_rate": 4.873010837959292e-05,
+      "loss": 1.4323,
+      "step": 4814
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8994624614715576,
+      "learning_rate": 4.872984403912239e-05,
+      "loss": 1.4792,
+      "step": 4815
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.155423879623413,
+      "learning_rate": 4.872957969865187e-05,
+      "loss": 1.4436,
+      "step": 4816
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.427677631378174,
+      "learning_rate": 4.8729315358181335e-05,
+      "loss": 2.2441,
+      "step": 4817
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.907751441001892,
+      "learning_rate": 4.8729051017710815e-05,
+      "loss": 1.4547,
+      "step": 4818
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.598409652709961,
+      "learning_rate": 4.872878667724029e-05,
+      "loss": 1.2445,
+      "step": 4819
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4925880432128906,
+      "learning_rate": 4.872852233676976e-05,
+      "loss": 1.3039,
+      "step": 4820
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7249337434768677,
+      "learning_rate": 4.8728257996299236e-05,
+      "loss": 1.3588,
+      "step": 4821
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5256850719451904,
+      "learning_rate": 4.872799365582871e-05,
+      "loss": 2.3031,
+      "step": 4822
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6662389039993286,
+      "learning_rate": 4.872772931535818e-05,
+      "loss": 1.4293,
+      "step": 4823
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5917589664459229,
+      "learning_rate": 4.872746497488766e-05,
+      "loss": 1.4059,
+      "step": 4824
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7181929349899292,
+      "learning_rate": 4.872720063441713e-05,
+      "loss": 1.4752,
+      "step": 4825
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.103269577026367,
+      "learning_rate": 4.8726936293946604e-05,
+      "loss": 1.5816,
+      "step": 4826
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8674312829971313,
+      "learning_rate": 4.872667195347608e-05,
+      "loss": 1.5292,
+      "step": 4827
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5836228132247925,
+      "learning_rate": 4.872640761300555e-05,
+      "loss": 1.3905,
+      "step": 4828
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7339340448379517,
+      "learning_rate": 4.872614327253503e-05,
+      "loss": 1.2905,
+      "step": 4829
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6918506622314453,
+      "learning_rate": 4.87258789320645e-05,
+      "loss": 1.3638,
+      "step": 4830
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.435903549194336,
+      "learning_rate": 4.872561459159398e-05,
+      "loss": 1.3413,
+      "step": 4831
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.754869818687439,
+      "learning_rate": 4.8725350251123445e-05,
+      "loss": 1.4561,
+      "step": 4832
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4757797718048096,
+      "learning_rate": 4.8725085910652926e-05,
+      "loss": 1.8437,
+      "step": 4833
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1193881034851074,
+      "learning_rate": 4.87248215701824e-05,
+      "loss": 1.3421,
+      "step": 4834
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6901538372039795,
+      "learning_rate": 4.872455722971187e-05,
+      "loss": 1.4342,
+      "step": 4835
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5628629922866821,
+      "learning_rate": 4.8724292889241346e-05,
+      "loss": 1.767,
+      "step": 4836
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.030682325363159,
+      "learning_rate": 4.872402854877082e-05,
+      "loss": 1.5656,
+      "step": 4837
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6896750926971436,
+      "learning_rate": 4.872376420830029e-05,
+      "loss": 1.4125,
+      "step": 4838
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6047570705413818,
+      "learning_rate": 4.872349986782977e-05,
+      "loss": 1.2692,
+      "step": 4839
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5963072776794434,
+      "learning_rate": 4.872323552735924e-05,
+      "loss": 1.2769,
+      "step": 4840
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8427115678787231,
+      "learning_rate": 4.8722971186888714e-05,
+      "loss": 1.5022,
+      "step": 4841
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4171395301818848,
+      "learning_rate": 4.872270684641819e-05,
+      "loss": 1.2721,
+      "step": 4842
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.891497015953064,
+      "learning_rate": 4.872244250594766e-05,
+      "loss": 1.4584,
+      "step": 4843
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2685813903808594,
+      "learning_rate": 4.872217816547714e-05,
+      "loss": 1.4546,
+      "step": 4844
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5517158508300781,
+      "learning_rate": 4.872191382500661e-05,
+      "loss": 1.3149,
+      "step": 4845
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9904347658157349,
+      "learning_rate": 4.872164948453609e-05,
+      "loss": 1.4788,
+      "step": 4846
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1975393295288086,
+      "learning_rate": 4.8721385144065556e-05,
+      "loss": 1.6075,
+      "step": 4847
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7787474393844604,
+      "learning_rate": 4.8721120803595036e-05,
+      "loss": 1.3231,
+      "step": 4848
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4164040088653564,
+      "learning_rate": 4.872085646312451e-05,
+      "loss": 1.4263,
+      "step": 4849
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5439420938491821,
+      "learning_rate": 4.872059212265398e-05,
+      "loss": 1.4663,
+      "step": 4850
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9263813495635986,
+      "learning_rate": 4.8720327782183457e-05,
+      "loss": 1.3349,
+      "step": 4851
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0981009006500244,
+      "learning_rate": 4.872006344171292e-05,
+      "loss": 1.5172,
+      "step": 4852
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9816277027130127,
+      "learning_rate": 4.8719799101242404e-05,
+      "loss": 1.3621,
+      "step": 4853
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1633105278015137,
+      "learning_rate": 4.871953476077188e-05,
+      "loss": 1.4803,
+      "step": 4854
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6275383234024048,
+      "learning_rate": 4.871927042030135e-05,
+      "loss": 1.2384,
+      "step": 4855
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.741591453552246,
+      "learning_rate": 4.8719006079830824e-05,
+      "loss": 1.3224,
+      "step": 4856
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6078968048095703,
+      "learning_rate": 4.87187417393603e-05,
+      "loss": 1.3135,
+      "step": 4857
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.857398271560669,
+      "learning_rate": 4.871847739888977e-05,
+      "loss": 1.4416,
+      "step": 4858
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.09887957572937,
+      "learning_rate": 4.871821305841925e-05,
+      "loss": 1.443,
+      "step": 4859
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4334745407104492,
+      "learning_rate": 4.871794871794872e-05,
+      "loss": 2.434,
+      "step": 4860
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7403135299682617,
+      "learning_rate": 4.871768437747819e-05,
+      "loss": 1.3502,
+      "step": 4861
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7125118970870972,
+      "learning_rate": 4.8717420037007666e-05,
+      "loss": 1.3211,
+      "step": 4862
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.42034113407135,
+      "learning_rate": 4.871715569653714e-05,
+      "loss": 1.4056,
+      "step": 4863
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7776933908462524,
+      "learning_rate": 4.871689135606662e-05,
+      "loss": 1.443,
+      "step": 4864
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1161253452301025,
+      "learning_rate": 4.8716627015596087e-05,
+      "loss": 1.4796,
+      "step": 4865
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8451485633850098,
+      "learning_rate": 4.871636267512557e-05,
+      "loss": 1.3786,
+      "step": 4866
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6981666088104248,
+      "learning_rate": 4.8716098334655034e-05,
+      "loss": 1.3027,
+      "step": 4867
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3414359092712402,
+      "learning_rate": 4.8715833994184514e-05,
+      "loss": 1.6477,
+      "step": 4868
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5654170513153076,
+      "learning_rate": 4.871556965371399e-05,
+      "loss": 1.3826,
+      "step": 4869
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.914113163948059,
+      "learning_rate": 4.871530531324346e-05,
+      "loss": 1.5188,
+      "step": 4870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.752875566482544,
+      "learning_rate": 4.8715040972772935e-05,
+      "loss": 1.4261,
+      "step": 4871
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6675935983657837,
+      "learning_rate": 4.871477663230241e-05,
+      "loss": 1.4167,
+      "step": 4872
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.054558038711548,
+      "learning_rate": 4.871451229183188e-05,
+      "loss": 1.4098,
+      "step": 4873
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7884800434112549,
+      "learning_rate": 4.8714247951361355e-05,
+      "loss": 1.3988,
+      "step": 4874
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9767447710037231,
+      "learning_rate": 4.871398361089083e-05,
+      "loss": 1.3362,
+      "step": 4875
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.830731749534607,
+      "learning_rate": 4.87137192704203e-05,
+      "loss": 1.561,
+      "step": 4876
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9263410568237305,
+      "learning_rate": 4.8713454929949776e-05,
+      "loss": 1.3608,
+      "step": 4877
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8818109035491943,
+      "learning_rate": 4.871319058947925e-05,
+      "loss": 1.4381,
+      "step": 4878
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7379997968673706,
+      "learning_rate": 4.871292624900873e-05,
+      "loss": 1.4269,
+      "step": 4879
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3397257328033447,
+      "learning_rate": 4.87126619085382e-05,
+      "loss": 1.3294,
+      "step": 4880
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7514376640319824,
+      "learning_rate": 4.871239756806768e-05,
+      "loss": 1.5301,
+      "step": 4881
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1751139163970947,
+      "learning_rate": 4.8712133227597144e-05,
+      "loss": 1.3862,
+      "step": 4882
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.22060489654541,
+      "learning_rate": 4.8711868887126624e-05,
+      "loss": 1.5726,
+      "step": 4883
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.124983549118042,
+      "learning_rate": 4.87116045466561e-05,
+      "loss": 1.7173,
+      "step": 4884
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8005260229110718,
+      "learning_rate": 4.871134020618557e-05,
+      "loss": 1.5724,
+      "step": 4885
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3444573879241943,
+      "learning_rate": 4.8711075865715045e-05,
+      "loss": 1.1447,
+      "step": 4886
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5897128582000732,
+      "learning_rate": 4.871081152524451e-05,
+      "loss": 1.4529,
+      "step": 4887
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0316367149353027,
+      "learning_rate": 4.871054718477399e-05,
+      "loss": 1.5668,
+      "step": 4888
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.887563943862915,
+      "learning_rate": 4.8710282844303466e-05,
+      "loss": 1.5522,
+      "step": 4889
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.261098861694336,
+      "learning_rate": 4.871001850383294e-05,
+      "loss": 1.4841,
+      "step": 4890
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4546172618865967,
+      "learning_rate": 4.870975416336241e-05,
+      "loss": 1.6599,
+      "step": 4891
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4743642807006836,
+      "learning_rate": 4.8709489822891887e-05,
+      "loss": 1.2927,
+      "step": 4892
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7614916563034058,
+      "learning_rate": 4.870922548242136e-05,
+      "loss": 1.4091,
+      "step": 4893
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7554306983947754,
+      "learning_rate": 4.870896114195084e-05,
+      "loss": 1.411,
+      "step": 4894
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9833998680114746,
+      "learning_rate": 4.870869680148031e-05,
+      "loss": 1.3969,
+      "step": 4895
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9117248058319092,
+      "learning_rate": 4.870843246100978e-05,
+      "loss": 1.2844,
+      "step": 4896
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4015650749206543,
+      "learning_rate": 4.8708168120539254e-05,
+      "loss": 1.2772,
+      "step": 4897
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4771056175231934,
+      "learning_rate": 4.870790378006873e-05,
+      "loss": 1.4437,
+      "step": 4898
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4201390743255615,
+      "learning_rate": 4.870763943959821e-05,
+      "loss": 1.3214,
+      "step": 4899
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.420214056968689,
+      "learning_rate": 4.8707375099127675e-05,
+      "loss": 1.381,
+      "step": 4900
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6990818977355957,
+      "learning_rate": 4.8707110758657155e-05,
+      "loss": 1.244,
+      "step": 4901
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.626495599746704,
+      "learning_rate": 4.870684641818662e-05,
+      "loss": 1.398,
+      "step": 4902
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7529970407485962,
+      "learning_rate": 4.87065820777161e-05,
+      "loss": 1.391,
+      "step": 4903
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.916607141494751,
+      "learning_rate": 4.8706317737245576e-05,
+      "loss": 1.4148,
+      "step": 4904
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8111000061035156,
+      "learning_rate": 4.870605339677505e-05,
+      "loss": 1.5441,
+      "step": 4905
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6291018724441528,
+      "learning_rate": 4.870578905630452e-05,
+      "loss": 2.6488,
+      "step": 4906
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7076513767242432,
+      "learning_rate": 4.8705524715834e-05,
+      "loss": 1.4135,
+      "step": 4907
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7118021249771118,
+      "learning_rate": 4.870526037536347e-05,
+      "loss": 1.5969,
+      "step": 4908
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5311963558197021,
+      "learning_rate": 4.8704996034892944e-05,
+      "loss": 1.3756,
+      "step": 4909
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.625398635864258,
+      "learning_rate": 4.870473169442242e-05,
+      "loss": 1.3801,
+      "step": 4910
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.183727264404297,
+      "learning_rate": 4.870446735395189e-05,
+      "loss": 1.5007,
+      "step": 4911
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7470940351486206,
+      "learning_rate": 4.8704203013481365e-05,
+      "loss": 1.3476,
+      "step": 4912
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9258061647415161,
+      "learning_rate": 4.870393867301084e-05,
+      "loss": 2.5348,
+      "step": 4913
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.545506238937378,
+      "learning_rate": 4.870367433254032e-05,
+      "loss": 1.183,
+      "step": 4914
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3804837465286255,
+      "learning_rate": 4.8703409992069785e-05,
+      "loss": 1.3259,
+      "step": 4915
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.063900947570801,
+      "learning_rate": 4.8703145651599266e-05,
+      "loss": 1.5157,
+      "step": 4916
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3727660179138184,
+      "learning_rate": 4.870288131112873e-05,
+      "loss": 1.5529,
+      "step": 4917
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.781196117401123,
+      "learning_rate": 4.870261697065821e-05,
+      "loss": 1.2896,
+      "step": 4918
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6805131435394287,
+      "learning_rate": 4.8702352630187687e-05,
+      "loss": 1.4201,
+      "step": 4919
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8665285110473633,
+      "learning_rate": 4.870208828971716e-05,
+      "loss": 1.488,
+      "step": 4920
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6670989990234375,
+      "learning_rate": 4.8701823949246634e-05,
+      "loss": 1.581,
+      "step": 4921
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7260947227478027,
+      "learning_rate": 4.870155960877611e-05,
+      "loss": 1.5359,
+      "step": 4922
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7528116703033447,
+      "learning_rate": 4.870129526830558e-05,
+      "loss": 1.3278,
+      "step": 4923
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3746116161346436,
+      "learning_rate": 4.8701030927835054e-05,
+      "loss": 1.3858,
+      "step": 4924
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.162801504135132,
+      "learning_rate": 4.870076658736453e-05,
+      "loss": 1.4296,
+      "step": 4925
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.39121150970459,
+      "learning_rate": 4.8700502246894e-05,
+      "loss": 1.5256,
+      "step": 4926
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8415513038635254,
+      "learning_rate": 4.8700237906423475e-05,
+      "loss": 3.0087,
+      "step": 4927
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4654817581176758,
+      "learning_rate": 4.869997356595295e-05,
+      "loss": 1.3305,
+      "step": 4928
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3570349216461182,
+      "learning_rate": 4.869970922548243e-05,
+      "loss": 1.2556,
+      "step": 4929
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3997691869735718,
+      "learning_rate": 4.8699444885011896e-05,
+      "loss": 1.3417,
+      "step": 4930
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5318349599838257,
+      "learning_rate": 4.8699180544541376e-05,
+      "loss": 2.0942,
+      "step": 4931
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7220431566238403,
+      "learning_rate": 4.869891620407084e-05,
+      "loss": 1.2682,
+      "step": 4932
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8325737714767456,
+      "learning_rate": 4.8698651863600317e-05,
+      "loss": 1.3294,
+      "step": 4933
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3520660400390625,
+      "learning_rate": 4.86983875231298e-05,
+      "loss": 1.3696,
+      "step": 4934
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8604124784469604,
+      "learning_rate": 4.8698123182659264e-05,
+      "loss": 1.4117,
+      "step": 4935
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6230766773223877,
+      "learning_rate": 4.8697858842188744e-05,
+      "loss": 1.5162,
+      "step": 4936
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1417346000671387,
+      "learning_rate": 4.869759450171821e-05,
+      "loss": 1.5656,
+      "step": 4937
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8626620769500732,
+      "learning_rate": 4.869733016124769e-05,
+      "loss": 1.4429,
+      "step": 4938
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1346328258514404,
+      "learning_rate": 4.8697065820777165e-05,
+      "loss": 1.459,
+      "step": 4939
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6939042806625366,
+      "learning_rate": 4.869680148030664e-05,
+      "loss": 1.391,
+      "step": 4940
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.152578353881836,
+      "learning_rate": 4.869653713983611e-05,
+      "loss": 1.4644,
+      "step": 4941
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9885435104370117,
+      "learning_rate": 4.8696272799365585e-05,
+      "loss": 1.5044,
+      "step": 4942
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7559928894042969,
+      "learning_rate": 4.869600845889506e-05,
+      "loss": 1.7947,
+      "step": 4943
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.999366283416748,
+      "learning_rate": 4.869574411842453e-05,
+      "loss": 1.4373,
+      "step": 4944
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5340073108673096,
+      "learning_rate": 4.8695479777954006e-05,
+      "loss": 1.421,
+      "step": 4945
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.163839340209961,
+      "learning_rate": 4.869521543748348e-05,
+      "loss": 1.5684,
+      "step": 4946
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.990251064300537,
+      "learning_rate": 4.869495109701295e-05,
+      "loss": 1.5094,
+      "step": 4947
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.107069253921509,
+      "learning_rate": 4.869468675654243e-05,
+      "loss": 1.7124,
+      "step": 4948
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7101927995681763,
+      "learning_rate": 4.869442241607191e-05,
+      "loss": 1.414,
+      "step": 4949
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8881794214248657,
+      "learning_rate": 4.8694158075601374e-05,
+      "loss": 1.518,
+      "step": 4950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0846381187438965,
+      "learning_rate": 4.8693893735130854e-05,
+      "loss": 1.6403,
+      "step": 4951
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4107463359832764,
+      "learning_rate": 4.869362939466032e-05,
+      "loss": 1.3371,
+      "step": 4952
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.329042673110962,
+      "learning_rate": 4.86933650541898e-05,
+      "loss": 1.2325,
+      "step": 4953
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6306116580963135,
+      "learning_rate": 4.8693100713719275e-05,
+      "loss": 1.3559,
+      "step": 4954
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0930492877960205,
+      "learning_rate": 4.869283637324875e-05,
+      "loss": 1.4891,
+      "step": 4955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6080923080444336,
+      "learning_rate": 4.869257203277822e-05,
+      "loss": 1.4454,
+      "step": 4956
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1149797439575195,
+      "learning_rate": 4.8692307692307696e-05,
+      "loss": 1.6517,
+      "step": 4957
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4818953275680542,
+      "learning_rate": 4.869204335183717e-05,
+      "loss": 1.4072,
+      "step": 4958
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9996486902236938,
+      "learning_rate": 4.869177901136664e-05,
+      "loss": 1.475,
+      "step": 4959
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.101475238800049,
+      "learning_rate": 4.8691514670896116e-05,
+      "loss": 1.4859,
+      "step": 4960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6621577739715576,
+      "learning_rate": 4.869125033042559e-05,
+      "loss": 1.6879,
+      "step": 4961
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5890787839889526,
+      "learning_rate": 4.8690985989955064e-05,
+      "loss": 1.4677,
+      "step": 4962
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.985701084136963,
+      "learning_rate": 4.869072164948454e-05,
+      "loss": 1.6159,
+      "step": 4963
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7709572315216064,
+      "learning_rate": 4.869045730901402e-05,
+      "loss": 1.385,
+      "step": 4964
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7723711729049683,
+      "learning_rate": 4.8690192968543484e-05,
+      "loss": 1.4397,
+      "step": 4965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.71100914478302,
+      "learning_rate": 4.8689928628072965e-05,
+      "loss": 1.4153,
+      "step": 4966
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2214081287384033,
+      "learning_rate": 4.868966428760243e-05,
+      "loss": 1.4886,
+      "step": 4967
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3692313432693481,
+      "learning_rate": 4.868939994713191e-05,
+      "loss": 1.4533,
+      "step": 4968
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8061403036117554,
+      "learning_rate": 4.8689135606661385e-05,
+      "loss": 1.3727,
+      "step": 4969
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5405685901641846,
+      "learning_rate": 4.868887126619085e-05,
+      "loss": 1.3286,
+      "step": 4970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7610901594161987,
+      "learning_rate": 4.868860692572033e-05,
+      "loss": 1.4285,
+      "step": 4971
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4152746200561523,
+      "learning_rate": 4.86883425852498e-05,
+      "loss": 1.34,
+      "step": 4972
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.18259334564209,
+      "learning_rate": 4.868807824477928e-05,
+      "loss": 1.4792,
+      "step": 4973
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5914876461029053,
+      "learning_rate": 4.868781390430875e-05,
+      "loss": 1.36,
+      "step": 4974
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.78889000415802,
+      "learning_rate": 4.868754956383823e-05,
+      "loss": 1.2635,
+      "step": 4975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6649785041809082,
+      "learning_rate": 4.86872852233677e-05,
+      "loss": 1.4446,
+      "step": 4976
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9915945529937744,
+      "learning_rate": 4.8687020882897174e-05,
+      "loss": 1.4059,
+      "step": 4977
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5898590087890625,
+      "learning_rate": 4.868675654242665e-05,
+      "loss": 1.3434,
+      "step": 4978
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9589594602584839,
+      "learning_rate": 4.868649220195612e-05,
+      "loss": 1.4204,
+      "step": 4979
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7360562086105347,
+      "learning_rate": 4.8686227861485595e-05,
+      "loss": 1.3124,
+      "step": 4980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7797917127609253,
+      "learning_rate": 4.868596352101507e-05,
+      "loss": 1.4867,
+      "step": 4981
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6740185022354126,
+      "learning_rate": 4.868569918054454e-05,
+      "loss": 1.2799,
+      "step": 4982
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5075536966323853,
+      "learning_rate": 4.8685434840074015e-05,
+      "loss": 1.3962,
+      "step": 4983
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.938774824142456,
+      "learning_rate": 4.8685170499603496e-05,
+      "loss": 1.4915,
+      "step": 4984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7410688400268555,
+      "learning_rate": 4.868490615913296e-05,
+      "loss": 2.0968,
+      "step": 4985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0958733558654785,
+      "learning_rate": 4.868464181866244e-05,
+      "loss": 1.5669,
+      "step": 4986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.537775993347168,
+      "learning_rate": 4.868437747819191e-05,
+      "loss": 1.3986,
+      "step": 4987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7573102712631226,
+      "learning_rate": 4.868411313772139e-05,
+      "loss": 1.3394,
+      "step": 4988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.996645450592041,
+      "learning_rate": 4.8683848797250864e-05,
+      "loss": 1.5028,
+      "step": 4989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7930316925048828,
+      "learning_rate": 4.868358445678034e-05,
+      "loss": 1.4475,
+      "step": 4990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1990103721618652,
+      "learning_rate": 4.868332011630981e-05,
+      "loss": 1.4999,
+      "step": 4991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0059127807617188,
+      "learning_rate": 4.8683055775839284e-05,
+      "loss": 1.5409,
+      "step": 4992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.791382074356079,
+      "learning_rate": 4.868279143536876e-05,
+      "loss": 1.3947,
+      "step": 4993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5882424116134644,
+      "learning_rate": 4.868252709489823e-05,
+      "loss": 1.5499,
+      "step": 4994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.866978645324707,
+      "learning_rate": 4.8682262754427705e-05,
+      "loss": 1.4043,
+      "step": 4995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.144052267074585,
+      "learning_rate": 4.868199841395718e-05,
+      "loss": 1.4575,
+      "step": 4996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.118326187133789,
+      "learning_rate": 4.868173407348665e-05,
+      "loss": 1.4724,
+      "step": 4997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8918907642364502,
+      "learning_rate": 4.8681469733016126e-05,
+      "loss": 1.237,
+      "step": 4998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.443967580795288,
+      "learning_rate": 4.8681205392545606e-05,
+      "loss": 1.6475,
+      "step": 4999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8268206119537354,
+      "learning_rate": 4.868094105207507e-05,
+      "loss": 1.4047,
+      "step": 5000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.24735689163208,
+      "learning_rate": 4.868067671160455e-05,
+      "loss": 1.5096,
+      "step": 5001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7260456085205078,
+      "learning_rate": 4.868041237113402e-05,
+      "loss": 1.3706,
+      "step": 5002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4747235774993896,
+      "learning_rate": 4.86801480306635e-05,
+      "loss": 1.3047,
+      "step": 5003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2370574474334717,
+      "learning_rate": 4.8679883690192974e-05,
+      "loss": 1.5743,
+      "step": 5004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.809115171432495,
+      "learning_rate": 4.867961934972244e-05,
+      "loss": 1.5531,
+      "step": 5005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6601066589355469,
+      "learning_rate": 4.867935500925192e-05,
+      "loss": 1.4895,
+      "step": 5006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4406843185424805,
+      "learning_rate": 4.867909066878139e-05,
+      "loss": 1.3527,
+      "step": 5007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4552210569381714,
+      "learning_rate": 4.867882632831087e-05,
+      "loss": 1.4787,
+      "step": 5008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7379496097564697,
+      "learning_rate": 4.867856198784034e-05,
+      "loss": 1.4003,
+      "step": 5009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6198465824127197,
+      "learning_rate": 4.8678297647369815e-05,
+      "loss": 1.3674,
+      "step": 5010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0635592937469482,
+      "learning_rate": 4.867803330689929e-05,
+      "loss": 1.5495,
+      "step": 5011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.642282247543335,
+      "learning_rate": 4.867776896642876e-05,
+      "loss": 1.2642,
+      "step": 5012
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.702661395072937,
+      "learning_rate": 4.8677504625958236e-05,
+      "loss": 1.4115,
+      "step": 5013
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.790982723236084,
+      "learning_rate": 4.867724028548771e-05,
+      "loss": 1.4106,
+      "step": 5014
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3296934366226196,
+      "learning_rate": 4.867697594501718e-05,
+      "loss": 1.3168,
+      "step": 5015
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8701715469360352,
+      "learning_rate": 4.867671160454666e-05,
+      "loss": 1.5088,
+      "step": 5016
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7966537475585938,
+      "learning_rate": 4.867644726407613e-05,
+      "loss": 1.3776,
+      "step": 5017
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9877312183380127,
+      "learning_rate": 4.8676182923605604e-05,
+      "loss": 1.4264,
+      "step": 5018
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.591387152671814,
+      "learning_rate": 4.8675918583135084e-05,
+      "loss": 1.5031,
+      "step": 5019
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5420416593551636,
+      "learning_rate": 4.867565424266455e-05,
+      "loss": 1.3927,
+      "step": 5020
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4143332242965698,
+      "learning_rate": 4.867538990219403e-05,
+      "loss": 1.3418,
+      "step": 5021
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8801381587982178,
+      "learning_rate": 4.86751255617235e-05,
+      "loss": 1.476,
+      "step": 5022
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6478792428970337,
+      "learning_rate": 4.867486122125298e-05,
+      "loss": 1.3543,
+      "step": 5023
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.222301721572876,
+      "learning_rate": 4.867459688078245e-05,
+      "loss": 1.5316,
+      "step": 5024
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.111440420150757,
+      "learning_rate": 4.8674332540311926e-05,
+      "loss": 1.4881,
+      "step": 5025
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.736323595046997,
+      "learning_rate": 4.86740681998414e-05,
+      "loss": 1.488,
+      "step": 5026
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1206703186035156,
+      "learning_rate": 4.867380385937087e-05,
+      "loss": 1.3481,
+      "step": 5027
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6382393836975098,
+      "learning_rate": 4.8673539518900346e-05,
+      "loss": 1.4482,
+      "step": 5028
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5581668615341187,
+      "learning_rate": 4.867327517842982e-05,
+      "loss": 1.2957,
+      "step": 5029
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4325802326202393,
+      "learning_rate": 4.8673010837959294e-05,
+      "loss": 1.4865,
+      "step": 5030
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7588595151901245,
+      "learning_rate": 4.867274649748877e-05,
+      "loss": 1.326,
+      "step": 5031
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.841228723526001,
+      "learning_rate": 4.867248215701824e-05,
+      "loss": 1.3451,
+      "step": 5032
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2284538745880127,
+      "learning_rate": 4.8672217816547714e-05,
+      "loss": 1.3986,
+      "step": 5033
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9486496448516846,
+      "learning_rate": 4.8671953476077195e-05,
+      "loss": 1.4235,
+      "step": 5034
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4527573585510254,
+      "learning_rate": 4.867168913560666e-05,
+      "loss": 1.3647,
+      "step": 5035
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.705467700958252,
+      "learning_rate": 4.867142479513614e-05,
+      "loss": 1.3936,
+      "step": 5036
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.141799211502075,
+      "learning_rate": 4.867116045466561e-05,
+      "loss": 1.4333,
+      "step": 5037
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6164681911468506,
+      "learning_rate": 4.867089611419509e-05,
+      "loss": 1.3377,
+      "step": 5038
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.360731840133667,
+      "learning_rate": 4.867063177372456e-05,
+      "loss": 1.3415,
+      "step": 5039
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9524494409561157,
+      "learning_rate": 4.8670367433254036e-05,
+      "loss": 1.4123,
+      "step": 5040
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9822616577148438,
+      "learning_rate": 4.867010309278351e-05,
+      "loss": 1.4006,
+      "step": 5041
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5218133926391602,
+      "learning_rate": 4.8669838752312976e-05,
+      "loss": 1.3602,
+      "step": 5042
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.862813115119934,
+      "learning_rate": 4.866957441184246e-05,
+      "loss": 1.2894,
+      "step": 5043
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.867576003074646,
+      "learning_rate": 4.866931007137193e-05,
+      "loss": 1.4351,
+      "step": 5044
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5421282052993774,
+      "learning_rate": 4.8669045730901404e-05,
+      "loss": 1.2438,
+      "step": 5045
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.837652564048767,
+      "learning_rate": 4.866878139043088e-05,
+      "loss": 1.406,
+      "step": 5046
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7469404935836792,
+      "learning_rate": 4.866851704996035e-05,
+      "loss": 1.3825,
+      "step": 5047
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6525826454162598,
+      "learning_rate": 4.8668252709489825e-05,
+      "loss": 2.6353,
+      "step": 5048
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6018587350845337,
+      "learning_rate": 4.8667988369019305e-05,
+      "loss": 1.3771,
+      "step": 5049
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0618960857391357,
+      "learning_rate": 4.866772402854877e-05,
+      "loss": 1.5362,
+      "step": 5050
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4860055446624756,
+      "learning_rate": 4.8667459688078245e-05,
+      "loss": 1.4475,
+      "step": 5051
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9671741724014282,
+      "learning_rate": 4.866719534760772e-05,
+      "loss": 1.3527,
+      "step": 5052
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5913745164871216,
+      "learning_rate": 4.866693100713719e-05,
+      "loss": 1.3251,
+      "step": 5053
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8517831563949585,
+      "learning_rate": 4.866666666666667e-05,
+      "loss": 1.4363,
+      "step": 5054
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7905502319335938,
+      "learning_rate": 4.866640232619614e-05,
+      "loss": 1.2657,
+      "step": 5055
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7300221920013428,
+      "learning_rate": 4.866613798572562e-05,
+      "loss": 1.6431,
+      "step": 5056
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7086982727050781,
+      "learning_rate": 4.866587364525509e-05,
+      "loss": 1.3983,
+      "step": 5057
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5340884923934937,
+      "learning_rate": 4.866560930478457e-05,
+      "loss": 1.3374,
+      "step": 5058
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.071272373199463,
+      "learning_rate": 4.866534496431404e-05,
+      "loss": 1.4829,
+      "step": 5059
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7352248430252075,
+      "learning_rate": 4.8665080623843514e-05,
+      "loss": 1.2553,
+      "step": 5060
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9067277908325195,
+      "learning_rate": 4.866481628337299e-05,
+      "loss": 1.3408,
+      "step": 5061
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.001325845718384,
+      "learning_rate": 4.866455194290246e-05,
+      "loss": 1.4967,
+      "step": 5062
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6023592948913574,
+      "learning_rate": 4.8664287602431935e-05,
+      "loss": 1.2613,
+      "step": 5063
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9130779504776,
+      "learning_rate": 4.866402326196141e-05,
+      "loss": 1.3248,
+      "step": 5064
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3408136367797852,
+      "learning_rate": 4.866375892149088e-05,
+      "loss": 1.1721,
+      "step": 5065
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.207916259765625,
+      "learning_rate": 4.8663494581020356e-05,
+      "loss": 1.3836,
+      "step": 5066
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.728011131286621,
+      "learning_rate": 4.866323024054983e-05,
+      "loss": 2.1709,
+      "step": 5067
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7476557493209839,
+      "learning_rate": 4.86629659000793e-05,
+      "loss": 1.1837,
+      "step": 5068
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.622424840927124,
+      "learning_rate": 4.866270155960878e-05,
+      "loss": 1.2697,
+      "step": 5069
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5526937246322632,
+      "learning_rate": 4.866243721913825e-05,
+      "loss": 1.422,
+      "step": 5070
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6929497718811035,
+      "learning_rate": 4.866217287866773e-05,
+      "loss": 1.4774,
+      "step": 5071
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.156050205230713,
+      "learning_rate": 4.86619085381972e-05,
+      "loss": 1.4024,
+      "step": 5072
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.189894437789917,
+      "learning_rate": 4.866164419772668e-05,
+      "loss": 1.323,
+      "step": 5073
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9643501043319702,
+      "learning_rate": 4.866137985725615e-05,
+      "loss": 1.3736,
+      "step": 5074
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6978962421417236,
+      "learning_rate": 4.8661115516785625e-05,
+      "loss": 1.3975,
+      "step": 5075
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9815614223480225,
+      "learning_rate": 4.86608511763151e-05,
+      "loss": 2.4238,
+      "step": 5076
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8065097332000732,
+      "learning_rate": 4.8660586835844565e-05,
+      "loss": 1.3392,
+      "step": 5077
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7947235107421875,
+      "learning_rate": 4.8660322495374045e-05,
+      "loss": 1.2982,
+      "step": 5078
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4946579933166504,
+      "learning_rate": 4.866005815490352e-05,
+      "loss": 1.4141,
+      "step": 5079
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7466166019439697,
+      "learning_rate": 4.865979381443299e-05,
+      "loss": 1.5489,
+      "step": 5080
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1630859375,
+      "learning_rate": 4.8659529473962466e-05,
+      "loss": 1.4188,
+      "step": 5081
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5530015230178833,
+      "learning_rate": 4.865926513349194e-05,
+      "loss": 1.4211,
+      "step": 5082
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6884865760803223,
+      "learning_rate": 4.865900079302141e-05,
+      "loss": 1.5224,
+      "step": 5083
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.116994619369507,
+      "learning_rate": 4.8658736452550893e-05,
+      "loss": 1.3956,
+      "step": 5084
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.806532382965088,
+      "learning_rate": 4.865847211208036e-05,
+      "loss": 1.4536,
+      "step": 5085
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.477004528045654,
+      "learning_rate": 4.865820777160984e-05,
+      "loss": 1.4803,
+      "step": 5086
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.375746011734009,
+      "learning_rate": 4.865794343113931e-05,
+      "loss": 1.6166,
+      "step": 5087
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0250003337860107,
+      "learning_rate": 4.865767909066878e-05,
+      "loss": 1.6302,
+      "step": 5088
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6378000974655151,
+      "learning_rate": 4.865741475019826e-05,
+      "loss": 1.3834,
+      "step": 5089
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7504799365997314,
+      "learning_rate": 4.865715040972773e-05,
+      "loss": 1.3927,
+      "step": 5090
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9740097522735596,
+      "learning_rate": 4.865688606925721e-05,
+      "loss": 1.3917,
+      "step": 5091
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9461942911148071,
+      "learning_rate": 4.8656621728786675e-05,
+      "loss": 1.4387,
+      "step": 5092
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7324118614196777,
+      "learning_rate": 4.8656357388316156e-05,
+      "loss": 1.4922,
+      "step": 5093
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5107766389846802,
+      "learning_rate": 4.865609304784563e-05,
+      "loss": 1.2631,
+      "step": 5094
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4451884031295776,
+      "learning_rate": 4.86558287073751e-05,
+      "loss": 1.4242,
+      "step": 5095
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6173384189605713,
+      "learning_rate": 4.8655564366904576e-05,
+      "loss": 1.44,
+      "step": 5096
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.003018379211426,
+      "learning_rate": 4.865530002643405e-05,
+      "loss": 1.543,
+      "step": 5097
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1760337352752686,
+      "learning_rate": 4.8655035685963523e-05,
+      "loss": 1.5796,
+      "step": 5098
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4177656173706055,
+      "learning_rate": 4.8654771345493e-05,
+      "loss": 1.4138,
+      "step": 5099
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5772087574005127,
+      "learning_rate": 4.865450700502247e-05,
+      "loss": 2.5194,
+      "step": 5100
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6280790567398071,
+      "learning_rate": 4.8654242664551944e-05,
+      "loss": 1.4554,
+      "step": 5101
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7518951892852783,
+      "learning_rate": 4.865397832408142e-05,
+      "loss": 1.2546,
+      "step": 5102
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.686144471168518,
+      "learning_rate": 4.865371398361089e-05,
+      "loss": 1.4888,
+      "step": 5103
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8135077953338623,
+      "learning_rate": 4.865344964314037e-05,
+      "loss": 1.5054,
+      "step": 5104
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0555739402770996,
+      "learning_rate": 4.865318530266984e-05,
+      "loss": 1.5204,
+      "step": 5105
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9114511013031006,
+      "learning_rate": 4.865292096219932e-05,
+      "loss": 1.6313,
+      "step": 5106
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.143394947052002,
+      "learning_rate": 4.8652656621728786e-05,
+      "loss": 1.4484,
+      "step": 5107
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.354278802871704,
+      "learning_rate": 4.8652392281258266e-05,
+      "loss": 1.5967,
+      "step": 5108
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.132824420928955,
+      "learning_rate": 4.865212794078774e-05,
+      "loss": 1.3833,
+      "step": 5109
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2990378141403198,
+      "learning_rate": 4.865186360031721e-05,
+      "loss": 1.2668,
+      "step": 5110
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6100178956985474,
+      "learning_rate": 4.865159925984669e-05,
+      "loss": 1.2112,
+      "step": 5111
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.197657823562622,
+      "learning_rate": 4.865133491937616e-05,
+      "loss": 1.5034,
+      "step": 5112
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5683358907699585,
+      "learning_rate": 4.8651070578905634e-05,
+      "loss": 1.3776,
+      "step": 5113
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5644341707229614,
+      "learning_rate": 4.865080623843511e-05,
+      "loss": 1.3893,
+      "step": 5114
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7618128061294556,
+      "learning_rate": 4.865054189796458e-05,
+      "loss": 1.5122,
+      "step": 5115
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0988450050354004,
+      "learning_rate": 4.8650277557494055e-05,
+      "loss": 1.4672,
+      "step": 5116
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.962446689605713,
+      "learning_rate": 4.865001321702353e-05,
+      "loss": 1.39,
+      "step": 5117
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3046679496765137,
+      "learning_rate": 4.8649748876553e-05,
+      "loss": 1.3446,
+      "step": 5118
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5809684991836548,
+      "learning_rate": 4.864948453608248e-05,
+      "loss": 1.2943,
+      "step": 5119
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5023223161697388,
+      "learning_rate": 4.864922019561195e-05,
+      "loss": 1.3917,
+      "step": 5120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6966707706451416,
+      "learning_rate": 4.864895585514143e-05,
+      "loss": 1.4311,
+      "step": 5121
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.817119836807251,
+      "learning_rate": 4.8648691514670896e-05,
+      "loss": 1.3295,
+      "step": 5122
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.026289463043213,
+      "learning_rate": 4.864842717420037e-05,
+      "loss": 1.4527,
+      "step": 5123
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.154320240020752,
+      "learning_rate": 4.864816283372985e-05,
+      "loss": 1.322,
+      "step": 5124
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3451181650161743,
+      "learning_rate": 4.864789849325932e-05,
+      "loss": 1.2704,
+      "step": 5125
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.708475947380066,
+      "learning_rate": 4.86476341527888e-05,
+      "loss": 1.4989,
+      "step": 5126
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6647648811340332,
+      "learning_rate": 4.8647369812318264e-05,
+      "loss": 1.2294,
+      "step": 5127
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.889760971069336,
+      "learning_rate": 4.8647105471847744e-05,
+      "loss": 1.6541,
+      "step": 5128
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.98449444770813,
+      "learning_rate": 4.864684113137722e-05,
+      "loss": 1.4252,
+      "step": 5129
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9654237031936646,
+      "learning_rate": 4.864657679090669e-05,
+      "loss": 1.3558,
+      "step": 5130
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9404886960983276,
+      "learning_rate": 4.8646312450436165e-05,
+      "loss": 1.3745,
+      "step": 5131
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9549217224121094,
+      "learning_rate": 4.864604810996564e-05,
+      "loss": 1.4589,
+      "step": 5132
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0656158924102783,
+      "learning_rate": 4.864578376949511e-05,
+      "loss": 1.4026,
+      "step": 5133
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.691159725189209,
+      "learning_rate": 4.8645519429024586e-05,
+      "loss": 1.4343,
+      "step": 5134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9129382371902466,
+      "learning_rate": 4.864525508855406e-05,
+      "loss": 1.4571,
+      "step": 5135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.16690731048584,
+      "learning_rate": 4.864499074808353e-05,
+      "loss": 1.3657,
+      "step": 5136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.461301565170288,
+      "learning_rate": 4.8644726407613006e-05,
+      "loss": 1.2724,
+      "step": 5137
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7162954807281494,
+      "learning_rate": 4.864446206714248e-05,
+      "loss": 1.3554,
+      "step": 5138
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.696945071220398,
+      "learning_rate": 4.864419772667196e-05,
+      "loss": 2.3007,
+      "step": 5139
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8093652725219727,
+      "learning_rate": 4.864393338620143e-05,
+      "loss": 1.3859,
+      "step": 5140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.279399871826172,
+      "learning_rate": 4.864366904573091e-05,
+      "loss": 1.3897,
+      "step": 5141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.997232437133789,
+      "learning_rate": 4.8643404705260374e-05,
+      "loss": 1.4645,
+      "step": 5142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.171142339706421,
+      "learning_rate": 4.8643140364789854e-05,
+      "loss": 1.4962,
+      "step": 5143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7879996299743652,
+      "learning_rate": 4.864287602431933e-05,
+      "loss": 1.272,
+      "step": 5144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6364942789077759,
+      "learning_rate": 4.86426116838488e-05,
+      "loss": 1.3649,
+      "step": 5145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1427197456359863,
+      "learning_rate": 4.8642347343378275e-05,
+      "loss": 1.5851,
+      "step": 5146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3754429817199707,
+      "learning_rate": 4.864208300290775e-05,
+      "loss": 1.4423,
+      "step": 5147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6472619771957397,
+      "learning_rate": 4.864181866243722e-05,
+      "loss": 1.4223,
+      "step": 5148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6785720586776733,
+      "learning_rate": 4.8641554321966696e-05,
+      "loss": 1.3706,
+      "step": 5149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5690518617630005,
+      "learning_rate": 4.864128998149617e-05,
+      "loss": 2.4114,
+      "step": 5150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6232547760009766,
+      "learning_rate": 4.864102564102564e-05,
+      "loss": 1.3078,
+      "step": 5151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5793025493621826,
+      "learning_rate": 4.864076130055512e-05,
+      "loss": 1.3698,
+      "step": 5152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8714380264282227,
+      "learning_rate": 4.864049696008459e-05,
+      "loss": 1.5237,
+      "step": 5153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.278896689414978,
+      "learning_rate": 4.864023261961407e-05,
+      "loss": 1.3948,
+      "step": 5154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.411090612411499,
+      "learning_rate": 4.863996827914354e-05,
+      "loss": 1.3068,
+      "step": 5155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7044093608856201,
+      "learning_rate": 4.863970393867302e-05,
+      "loss": 1.5902,
+      "step": 5156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8125293254852295,
+      "learning_rate": 4.8639439598202484e-05,
+      "loss": 1.465,
+      "step": 5157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.721221923828125,
+      "learning_rate": 4.8639175257731965e-05,
+      "loss": 1.4879,
+      "step": 5158
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0391621589660645,
+      "learning_rate": 4.863891091726144e-05,
+      "loss": 1.4802,
+      "step": 5159
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.469994306564331,
+      "learning_rate": 4.8638646576790905e-05,
+      "loss": 1.6064,
+      "step": 5160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.647261381149292,
+      "learning_rate": 4.8638382236320386e-05,
+      "loss": 1.3032,
+      "step": 5161
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7059807777404785,
+      "learning_rate": 4.863811789584985e-05,
+      "loss": 1.4779,
+      "step": 5162
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0523598194122314,
+      "learning_rate": 4.863785355537933e-05,
+      "loss": 1.6941,
+      "step": 5163
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.254281520843506,
+      "learning_rate": 4.8637589214908806e-05,
+      "loss": 1.5537,
+      "step": 5164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9392704963684082,
+      "learning_rate": 4.863732487443828e-05,
+      "loss": 1.4864,
+      "step": 5165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8917629718780518,
+      "learning_rate": 4.8637060533967753e-05,
+      "loss": 1.4737,
+      "step": 5166
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1095082759857178,
+      "learning_rate": 4.863679619349723e-05,
+      "loss": 1.3495,
+      "step": 5167
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.441009044647217,
+      "learning_rate": 4.86365318530267e-05,
+      "loss": 1.5119,
+      "step": 5168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1319315433502197,
+      "learning_rate": 4.8636267512556174e-05,
+      "loss": 1.371,
+      "step": 5169
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9224469661712646,
+      "learning_rate": 4.863600317208565e-05,
+      "loss": 1.4337,
+      "step": 5170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9686206579208374,
+      "learning_rate": 4.863573883161512e-05,
+      "loss": 1.4517,
+      "step": 5171
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6613092422485352,
+      "learning_rate": 4.8635474491144595e-05,
+      "loss": 1.4469,
+      "step": 5172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.084336519241333,
+      "learning_rate": 4.863521015067407e-05,
+      "loss": 1.611,
+      "step": 5173
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9563698768615723,
+      "learning_rate": 4.863494581020355e-05,
+      "loss": 1.4789,
+      "step": 5174
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.270129442214966,
+      "learning_rate": 4.8634681469733016e-05,
+      "loss": 1.6566,
+      "step": 5175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5717103481292725,
+      "learning_rate": 4.8634417129262496e-05,
+      "loss": 1.3256,
+      "step": 5176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.540058970451355,
+      "learning_rate": 4.863415278879196e-05,
+      "loss": 1.6302,
+      "step": 5177
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.580885887145996,
+      "learning_rate": 4.863388844832144e-05,
+      "loss": 1.4958,
+      "step": 5178
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.509296178817749,
+      "learning_rate": 4.8633624107850917e-05,
+      "loss": 1.273,
+      "step": 5179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.699367642402649,
+      "learning_rate": 4.863335976738039e-05,
+      "loss": 1.3618,
+      "step": 5180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6888995170593262,
+      "learning_rate": 4.8633095426909864e-05,
+      "loss": 1.3208,
+      "step": 5181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7175393104553223,
+      "learning_rate": 4.863283108643934e-05,
+      "loss": 1.5747,
+      "step": 5182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3608145713806152,
+      "learning_rate": 4.863256674596881e-05,
+      "loss": 1.5527,
+      "step": 5183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.977025032043457,
+      "learning_rate": 4.8632302405498284e-05,
+      "loss": 1.6915,
+      "step": 5184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.955486536026001,
+      "learning_rate": 4.863203806502776e-05,
+      "loss": 2.303,
+      "step": 5185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.615160346031189,
+      "learning_rate": 4.863177372455723e-05,
+      "loss": 1.4976,
+      "step": 5186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.901305079460144,
+      "learning_rate": 4.8631509384086705e-05,
+      "loss": 1.4898,
+      "step": 5187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.983221173286438,
+      "learning_rate": 4.863124504361618e-05,
+      "loss": 1.6239,
+      "step": 5188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5231177806854248,
+      "learning_rate": 4.863098070314566e-05,
+      "loss": 1.3807,
+      "step": 5189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5873147249221802,
+      "learning_rate": 4.8630716362675126e-05,
+      "loss": 1.4234,
+      "step": 5190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.704559564590454,
+      "learning_rate": 4.8630452022204606e-05,
+      "loss": 1.398,
+      "step": 5191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9087131023406982,
+      "learning_rate": 4.863018768173407e-05,
+      "loss": 1.5143,
+      "step": 5192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6564533710479736,
+      "learning_rate": 4.862992334126355e-05,
+      "loss": 1.415,
+      "step": 5193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.123513698577881,
+      "learning_rate": 4.862965900079303e-05,
+      "loss": 1.3254,
+      "step": 5194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.430173397064209,
+      "learning_rate": 4.8629394660322494e-05,
+      "loss": 1.3062,
+      "step": 5195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1393284797668457,
+      "learning_rate": 4.8629130319851974e-05,
+      "loss": 1.2798,
+      "step": 5196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.877774715423584,
+      "learning_rate": 4.862886597938144e-05,
+      "loss": 1.275,
+      "step": 5197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1143853664398193,
+      "learning_rate": 4.862860163891092e-05,
+      "loss": 1.3455,
+      "step": 5198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7126902341842651,
+      "learning_rate": 4.8628337298440395e-05,
+      "loss": 1.5206,
+      "step": 5199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5658602714538574,
+      "learning_rate": 4.862807295796987e-05,
+      "loss": 1.1394,
+      "step": 5200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7772653102874756,
+      "learning_rate": 4.862780861749934e-05,
+      "loss": 1.4499,
+      "step": 5201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9060330390930176,
+      "learning_rate": 4.8627544277028816e-05,
+      "loss": 1.407,
+      "step": 5202
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.212092161178589,
+      "learning_rate": 4.862727993655829e-05,
+      "loss": 1.5073,
+      "step": 5203
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2684082984924316,
+      "learning_rate": 4.862701559608776e-05,
+      "loss": 1.5551,
+      "step": 5204
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9860289096832275,
+      "learning_rate": 4.8626751255617236e-05,
+      "loss": 1.5618,
+      "step": 5205
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7210276126861572,
+      "learning_rate": 4.862648691514671e-05,
+      "loss": 1.2522,
+      "step": 5206
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8509480953216553,
+      "learning_rate": 4.862622257467618e-05,
+      "loss": 1.4695,
+      "step": 5207
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8806939125061035,
+      "learning_rate": 4.862595823420566e-05,
+      "loss": 1.5917,
+      "step": 5208
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2238192558288574,
+      "learning_rate": 4.862569389373514e-05,
+      "loss": 1.4199,
+      "step": 5209
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.701107144355774,
+      "learning_rate": 4.8625429553264604e-05,
+      "loss": 1.4004,
+      "step": 5210
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.935750126838684,
+      "learning_rate": 4.8625165212794084e-05,
+      "loss": 1.3025,
+      "step": 5211
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.043088912963867,
+      "learning_rate": 4.862490087232355e-05,
+      "loss": 1.4082,
+      "step": 5212
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6527788639068604,
+      "learning_rate": 4.862463653185303e-05,
+      "loss": 1.484,
+      "step": 5213
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4354079961776733,
+      "learning_rate": 4.8624372191382505e-05,
+      "loss": 2.3099,
+      "step": 5214
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7395429611206055,
+      "learning_rate": 4.862410785091198e-05,
+      "loss": 1.3817,
+      "step": 5215
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3159782886505127,
+      "learning_rate": 4.862384351044145e-05,
+      "loss": 1.2889,
+      "step": 5216
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.728847861289978,
+      "learning_rate": 4.8623579169970926e-05,
+      "loss": 1.487,
+      "step": 5217
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7162460088729858,
+      "learning_rate": 4.86233148295004e-05,
+      "loss": 1.4483,
+      "step": 5218
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4420044422149658,
+      "learning_rate": 4.862305048902987e-05,
+      "loss": 1.2062,
+      "step": 5219
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1409618854522705,
+      "learning_rate": 4.8622786148559347e-05,
+      "loss": 1.5862,
+      "step": 5220
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6815967559814453,
+      "learning_rate": 4.862252180808882e-05,
+      "loss": 1.6475,
+      "step": 5221
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.400904893875122,
+      "learning_rate": 4.8622257467618294e-05,
+      "loss": 1.3333,
+      "step": 5222
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7323731184005737,
+      "learning_rate": 4.862199312714777e-05,
+      "loss": 1.5423,
+      "step": 5223
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6293821334838867,
+      "learning_rate": 4.862172878667725e-05,
+      "loss": 1.4549,
+      "step": 5224
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5762555599212646,
+      "learning_rate": 4.8621464446206714e-05,
+      "loss": 1.2215,
+      "step": 5225
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5472909212112427,
+      "learning_rate": 4.8621200105736195e-05,
+      "loss": 1.1751,
+      "step": 5226
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2883856296539307,
+      "learning_rate": 4.862093576526566e-05,
+      "loss": 1.1654,
+      "step": 5227
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.78940486907959,
+      "learning_rate": 4.862067142479514e-05,
+      "loss": 1.3951,
+      "step": 5228
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6738594770431519,
+      "learning_rate": 4.8620407084324615e-05,
+      "loss": 1.2601,
+      "step": 5229
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1397128105163574,
+      "learning_rate": 4.862014274385409e-05,
+      "loss": 1.4699,
+      "step": 5230
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.526615023612976,
+      "learning_rate": 4.861987840338356e-05,
+      "loss": 1.3643,
+      "step": 5231
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5299111604690552,
+      "learning_rate": 4.861961406291303e-05,
+      "loss": 1.2837,
+      "step": 5232
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.182925224304199,
+      "learning_rate": 4.861934972244251e-05,
+      "loss": 1.4509,
+      "step": 5233
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6357624530792236,
+      "learning_rate": 4.861908538197198e-05,
+      "loss": 1.3796,
+      "step": 5234
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.882592797279358,
+      "learning_rate": 4.861882104150146e-05,
+      "loss": 1.3989,
+      "step": 5235
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6726585626602173,
+      "learning_rate": 4.861855670103093e-05,
+      "loss": 1.3531,
+      "step": 5236
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.534478783607483,
+      "learning_rate": 4.8618292360560404e-05,
+      "loss": 1.3926,
+      "step": 5237
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.859132289886475,
+      "learning_rate": 4.861802802008988e-05,
+      "loss": 1.6615,
+      "step": 5238
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3368959426879883,
+      "learning_rate": 4.861776367961936e-05,
+      "loss": 1.3818,
+      "step": 5239
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.780008316040039,
+      "learning_rate": 4.8617499339148825e-05,
+      "loss": 1.3415,
+      "step": 5240
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6896710395812988,
+      "learning_rate": 4.86172349986783e-05,
+      "loss": 1.4116,
+      "step": 5241
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.474116325378418,
+      "learning_rate": 4.861697065820777e-05,
+      "loss": 1.3624,
+      "step": 5242
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5540704727172852,
+      "learning_rate": 4.8616706317737245e-05,
+      "loss": 1.1933,
+      "step": 5243
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6778069734573364,
+      "learning_rate": 4.8616441977266726e-05,
+      "loss": 1.3281,
+      "step": 5244
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9394587278366089,
+      "learning_rate": 4.861617763679619e-05,
+      "loss": 1.4602,
+      "step": 5245
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8323668241500854,
+      "learning_rate": 4.861591329632567e-05,
+      "loss": 1.4203,
+      "step": 5246
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9505748748779297,
+      "learning_rate": 4.861564895585514e-05,
+      "loss": 1.487,
+      "step": 5247
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.699204683303833,
+      "learning_rate": 4.861538461538462e-05,
+      "loss": 1.2396,
+      "step": 5248
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.128918170928955,
+      "learning_rate": 4.8615120274914094e-05,
+      "loss": 1.4644,
+      "step": 5249
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.431837797164917,
+      "learning_rate": 4.861485593444357e-05,
+      "loss": 1.4867,
+      "step": 5250
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.304141044616699,
+      "learning_rate": 4.861459159397304e-05,
+      "loss": 1.4588,
+      "step": 5251
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9673115015029907,
+      "learning_rate": 4.8614327253502514e-05,
+      "loss": 1.5185,
+      "step": 5252
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5950638055801392,
+      "learning_rate": 4.861406291303199e-05,
+      "loss": 1.4012,
+      "step": 5253
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9103765487670898,
+      "learning_rate": 4.861379857256146e-05,
+      "loss": 1.4355,
+      "step": 5254
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0066678524017334,
+      "learning_rate": 4.8613534232090935e-05,
+      "loss": 1.5579,
+      "step": 5255
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2060670852661133,
+      "learning_rate": 4.861326989162041e-05,
+      "loss": 1.3533,
+      "step": 5256
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8741620779037476,
+      "learning_rate": 4.861300555114988e-05,
+      "loss": 1.4158,
+      "step": 5257
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5274467468261719,
+      "learning_rate": 4.8612741210679356e-05,
+      "loss": 1.2729,
+      "step": 5258
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7939972877502441,
+      "learning_rate": 4.8612476870208836e-05,
+      "loss": 1.4464,
+      "step": 5259
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7734266519546509,
+      "learning_rate": 4.86122125297383e-05,
+      "loss": 1.4242,
+      "step": 5260
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.499661922454834,
+      "learning_rate": 4.861194818926778e-05,
+      "loss": 1.1934,
+      "step": 5261
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.061842918395996,
+      "learning_rate": 4.861168384879725e-05,
+      "loss": 1.4154,
+      "step": 5262
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7919702529907227,
+      "learning_rate": 4.861141950832673e-05,
+      "loss": 1.494,
+      "step": 5263
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6177438497543335,
+      "learning_rate": 4.8611155167856204e-05,
+      "loss": 1.5063,
+      "step": 5264
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.259380340576172,
+      "learning_rate": 4.861089082738568e-05,
+      "loss": 1.4579,
+      "step": 5265
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.712118148803711,
+      "learning_rate": 4.861062648691515e-05,
+      "loss": 1.5319,
+      "step": 5266
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8486886024475098,
+      "learning_rate": 4.861036214644462e-05,
+      "loss": 1.3893,
+      "step": 5267
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4764600992202759,
+      "learning_rate": 4.86100978059741e-05,
+      "loss": 1.3526,
+      "step": 5268
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.481863021850586,
+      "learning_rate": 4.860983346550357e-05,
+      "loss": 1.2771,
+      "step": 5269
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0222883224487305,
+      "learning_rate": 4.8609569125033045e-05,
+      "loss": 1.4938,
+      "step": 5270
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.242063045501709,
+      "learning_rate": 4.860930478456252e-05,
+      "loss": 1.4664,
+      "step": 5271
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9941279888153076,
+      "learning_rate": 4.860904044409199e-05,
+      "loss": 1.4915,
+      "step": 5272
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2493793964385986,
+      "learning_rate": 4.8608776103621466e-05,
+      "loss": 1.385,
+      "step": 5273
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3426339626312256,
+      "learning_rate": 4.8608511763150946e-05,
+      "loss": 1.6404,
+      "step": 5274
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.509840250015259,
+      "learning_rate": 4.860824742268041e-05,
+      "loss": 1.4827,
+      "step": 5275
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5252368450164795,
+      "learning_rate": 4.8607983082209894e-05,
+      "loss": 1.3896,
+      "step": 5276
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9918278455734253,
+      "learning_rate": 4.860771874173936e-05,
+      "loss": 1.4212,
+      "step": 5277
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.295262575149536,
+      "learning_rate": 4.8607454401268834e-05,
+      "loss": 1.2787,
+      "step": 5278
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1779258251190186,
+      "learning_rate": 4.8607190060798314e-05,
+      "loss": 1.4167,
+      "step": 5279
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9088815450668335,
+      "learning_rate": 4.860692572032778e-05,
+      "loss": 1.4348,
+      "step": 5280
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6604983806610107,
+      "learning_rate": 4.860666137985726e-05,
+      "loss": 1.4857,
+      "step": 5281
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6753629446029663,
+      "learning_rate": 4.860639703938673e-05,
+      "loss": 1.248,
+      "step": 5282
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8594269752502441,
+      "learning_rate": 4.860613269891621e-05,
+      "loss": 1.3641,
+      "step": 5283
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.341987371444702,
+      "learning_rate": 4.860586835844568e-05,
+      "loss": 1.4702,
+      "step": 5284
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5739518404006958,
+      "learning_rate": 4.8605604017975156e-05,
+      "loss": 1.3778,
+      "step": 5285
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5440117120742798,
+      "learning_rate": 4.860533967750463e-05,
+      "loss": 1.3642,
+      "step": 5286
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9389474391937256,
+      "learning_rate": 4.86050753370341e-05,
+      "loss": 1.3904,
+      "step": 5287
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8202142715454102,
+      "learning_rate": 4.8604810996563576e-05,
+      "loss": 1.3607,
+      "step": 5288
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.357497453689575,
+      "learning_rate": 4.860454665609305e-05,
+      "loss": 1.4526,
+      "step": 5289
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.680070400238037,
+      "learning_rate": 4.8604282315622524e-05,
+      "loss": 1.3608,
+      "step": 5290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6809868812561035,
+      "learning_rate": 4.8604017975152e-05,
+      "loss": 1.4204,
+      "step": 5291
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6681894063949585,
+      "learning_rate": 4.860375363468147e-05,
+      "loss": 1.3506,
+      "step": 5292
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6051019430160522,
+      "learning_rate": 4.8603489294210944e-05,
+      "loss": 1.3064,
+      "step": 5293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4377756118774414,
+      "learning_rate": 4.8603224953740425e-05,
+      "loss": 1.2972,
+      "step": 5294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8569507598876953,
+      "learning_rate": 4.860296061326989e-05,
+      "loss": 1.5079,
+      "step": 5295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7837769985198975,
+      "learning_rate": 4.860269627279937e-05,
+      "loss": 1.3244,
+      "step": 5296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6019197702407837,
+      "learning_rate": 4.860243193232884e-05,
+      "loss": 1.2733,
+      "step": 5297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7037330865859985,
+      "learning_rate": 4.860216759185832e-05,
+      "loss": 1.363,
+      "step": 5298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1251752376556396,
+      "learning_rate": 4.860190325138779e-05,
+      "loss": 1.3945,
+      "step": 5299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.012937545776367,
+      "learning_rate": 4.8601638910917266e-05,
+      "loss": 1.4989,
+      "step": 5300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5322415828704834,
+      "learning_rate": 4.860137457044674e-05,
+      "loss": 1.4116,
+      "step": 5301
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0457534790039062,
+      "learning_rate": 4.860111022997621e-05,
+      "loss": 1.4438,
+      "step": 5302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7700626850128174,
+      "learning_rate": 4.860084588950569e-05,
+      "loss": 1.5939,
+      "step": 5303
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2125632762908936,
+      "learning_rate": 4.860058154903516e-05,
+      "loss": 1.624,
+      "step": 5304
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8780072927474976,
+      "learning_rate": 4.8600317208564634e-05,
+      "loss": 1.3636,
+      "step": 5305
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.618481159210205,
+      "learning_rate": 4.860005286809411e-05,
+      "loss": 1.3361,
+      "step": 5306
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6919130086898804,
+      "learning_rate": 4.859978852762358e-05,
+      "loss": 1.3672,
+      "step": 5307
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.613000750541687,
+      "learning_rate": 4.8599524187153055e-05,
+      "loss": 1.3711,
+      "step": 5308
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1275923252105713,
+      "learning_rate": 4.8599259846682535e-05,
+      "loss": 1.4044,
+      "step": 5309
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9263086318969727,
+      "learning_rate": 4.8598995506212e-05,
+      "loss": 1.3157,
+      "step": 5310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1215083599090576,
+      "learning_rate": 4.859873116574148e-05,
+      "loss": 1.4207,
+      "step": 5311
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.311781883239746,
+      "learning_rate": 4.859846682527095e-05,
+      "loss": 1.6229,
+      "step": 5312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6891224384307861,
+      "learning_rate": 4.859820248480042e-05,
+      "loss": 1.3518,
+      "step": 5313
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9205290079116821,
+      "learning_rate": 4.85979381443299e-05,
+      "loss": 1.4939,
+      "step": 5314
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8542691469192505,
+      "learning_rate": 4.859767380385937e-05,
+      "loss": 1.2753,
+      "step": 5315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.201920509338379,
+      "learning_rate": 4.859740946338885e-05,
+      "loss": 1.6075,
+      "step": 5316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.237804651260376,
+      "learning_rate": 4.859714512291832e-05,
+      "loss": 1.3216,
+      "step": 5317
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9954805374145508,
+      "learning_rate": 4.85968807824478e-05,
+      "loss": 1.3978,
+      "step": 5318
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.098306655883789,
+      "learning_rate": 4.859661644197727e-05,
+      "loss": 1.5868,
+      "step": 5319
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6422450542449951,
+      "learning_rate": 4.8596352101506744e-05,
+      "loss": 1.3257,
+      "step": 5320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8849369287490845,
+      "learning_rate": 4.859608776103622e-05,
+      "loss": 1.4423,
+      "step": 5321
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3586294651031494,
+      "learning_rate": 4.859582342056569e-05,
+      "loss": 1.2287,
+      "step": 5322
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7741249799728394,
+      "learning_rate": 4.8595559080095165e-05,
+      "loss": 1.3285,
+      "step": 5323
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6985574960708618,
+      "learning_rate": 4.859529473962464e-05,
+      "loss": 1.3417,
+      "step": 5324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.939887285232544,
+      "learning_rate": 4.859503039915411e-05,
+      "loss": 1.5941,
+      "step": 5325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8591536283493042,
+      "learning_rate": 4.8594766058683586e-05,
+      "loss": 1.5292,
+      "step": 5326
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8855643272399902,
+      "learning_rate": 4.859450171821306e-05,
+      "loss": 1.3384,
+      "step": 5327
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9386835098266602,
+      "learning_rate": 4.859423737774253e-05,
+      "loss": 1.4872,
+      "step": 5328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4981166124343872,
+      "learning_rate": 4.859397303727201e-05,
+      "loss": 1.1517,
+      "step": 5329
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6058554649353027,
+      "learning_rate": 4.859370869680148e-05,
+      "loss": 1.3916,
+      "step": 5330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6666791439056396,
+      "learning_rate": 4.859344435633096e-05,
+      "loss": 1.5627,
+      "step": 5331
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0114235877990723,
+      "learning_rate": 4.859318001586043e-05,
+      "loss": 1.445,
+      "step": 5332
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7128537893295288,
+      "learning_rate": 4.859291567538991e-05,
+      "loss": 1.3373,
+      "step": 5333
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4064152240753174,
+      "learning_rate": 4.8592651334919374e-05,
+      "loss": 1.3347,
+      "step": 5334
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.658583402633667,
+      "learning_rate": 4.8592386994448855e-05,
+      "loss": 1.3933,
+      "step": 5335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0252878665924072,
+      "learning_rate": 4.859212265397833e-05,
+      "loss": 1.4491,
+      "step": 5336
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.40341055393219,
+      "learning_rate": 4.85918583135078e-05,
+      "loss": 1.2962,
+      "step": 5337
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8543137311935425,
+      "learning_rate": 4.8591593973037275e-05,
+      "loss": 1.4765,
+      "step": 5338
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.006006956100464,
+      "learning_rate": 4.859132963256674e-05,
+      "loss": 1.4361,
+      "step": 5339
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8103505373001099,
+      "learning_rate": 4.859106529209622e-05,
+      "loss": 1.4489,
+      "step": 5340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4764364957809448,
+      "learning_rate": 4.8590800951625696e-05,
+      "loss": 1.3543,
+      "step": 5341
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5647832155227661,
+      "learning_rate": 4.859053661115517e-05,
+      "loss": 1.3379,
+      "step": 5342
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0770130157470703,
+      "learning_rate": 4.859027227068464e-05,
+      "loss": 1.5668,
+      "step": 5343
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8350965976715088,
+      "learning_rate": 4.859000793021412e-05,
+      "loss": 1.32,
+      "step": 5344
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4863150119781494,
+      "learning_rate": 4.858974358974359e-05,
+      "loss": 1.3547,
+      "step": 5345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0181949138641357,
+      "learning_rate": 4.858947924927307e-05,
+      "loss": 1.6005,
+      "step": 5346
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6453062295913696,
+      "learning_rate": 4.858921490880254e-05,
+      "loss": 1.2466,
+      "step": 5347
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.202205181121826,
+      "learning_rate": 4.858895056833202e-05,
+      "loss": 1.4206,
+      "step": 5348
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.192466974258423,
+      "learning_rate": 4.8588686227861485e-05,
+      "loss": 1.3864,
+      "step": 5349
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5104292631149292,
+      "learning_rate": 4.858842188739096e-05,
+      "loss": 1.4149,
+      "step": 5350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4650894403457642,
+      "learning_rate": 4.858815754692044e-05,
+      "loss": 1.3407,
+      "step": 5351
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.764079213142395,
+      "learning_rate": 4.8587893206449905e-05,
+      "loss": 1.4089,
+      "step": 5352
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7484022378921509,
+      "learning_rate": 4.8587628865979386e-05,
+      "loss": 2.3009,
+      "step": 5353
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7156037092208862,
+      "learning_rate": 4.858736452550885e-05,
+      "loss": 1.4585,
+      "step": 5354
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0516319274902344,
+      "learning_rate": 4.858710018503833e-05,
+      "loss": 1.5629,
+      "step": 5355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8973231315612793,
+      "learning_rate": 4.8586835844567806e-05,
+      "loss": 1.2199,
+      "step": 5356
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.481827974319458,
+      "learning_rate": 4.858657150409728e-05,
+      "loss": 1.3361,
+      "step": 5357
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5177333354949951,
+      "learning_rate": 4.8586307163626754e-05,
+      "loss": 1.4638,
+      "step": 5358
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7640187740325928,
+      "learning_rate": 4.858604282315623e-05,
+      "loss": 1.6318,
+      "step": 5359
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6444753408432007,
+      "learning_rate": 4.85857784826857e-05,
+      "loss": 1.5297,
+      "step": 5360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6124037504196167,
+      "learning_rate": 4.8585514142215174e-05,
+      "loss": 1.3351,
+      "step": 5361
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4506357908248901,
+      "learning_rate": 4.858524980174465e-05,
+      "loss": 1.3869,
+      "step": 5362
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.019822359085083,
+      "learning_rate": 4.858498546127412e-05,
+      "loss": 1.9224,
+      "step": 5363
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7862176895141602,
+      "learning_rate": 4.8584721120803595e-05,
+      "loss": 1.5377,
+      "step": 5364
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.13809871673584,
+      "learning_rate": 4.858445678033307e-05,
+      "loss": 1.5899,
+      "step": 5365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.552228331565857,
+      "learning_rate": 4.858419243986255e-05,
+      "loss": 1.4751,
+      "step": 5366
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0437185764312744,
+      "learning_rate": 4.8583928099392016e-05,
+      "loss": 1.4213,
+      "step": 5367
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7662135362625122,
+      "learning_rate": 4.8583663758921496e-05,
+      "loss": 1.5019,
+      "step": 5368
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6120871305465698,
+      "learning_rate": 4.858339941845096e-05,
+      "loss": 1.4858,
+      "step": 5369
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6122413873672485,
+      "learning_rate": 4.858313507798044e-05,
+      "loss": 1.4122,
+      "step": 5370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8890068531036377,
+      "learning_rate": 4.858287073750992e-05,
+      "loss": 1.4821,
+      "step": 5371
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8069190979003906,
+      "learning_rate": 4.858260639703939e-05,
+      "loss": 1.4151,
+      "step": 5372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.304452657699585,
+      "learning_rate": 4.8582342056568864e-05,
+      "loss": 1.336,
+      "step": 5373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.602577209472656,
+      "learning_rate": 4.858207771609834e-05,
+      "loss": 1.4936,
+      "step": 5374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8586231470108032,
+      "learning_rate": 4.858181337562781e-05,
+      "loss": 1.3953,
+      "step": 5375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5261013507843018,
+      "learning_rate": 4.8581549035157285e-05,
+      "loss": 1.352,
+      "step": 5376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.600139856338501,
+      "learning_rate": 4.858128469468676e-05,
+      "loss": 1.4158,
+      "step": 5377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.451575756072998,
+      "learning_rate": 4.858102035421623e-05,
+      "loss": 1.1993,
+      "step": 5378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2254111766815186,
+      "learning_rate": 4.8580756013745705e-05,
+      "loss": 1.5223,
+      "step": 5379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9579195976257324,
+      "learning_rate": 4.858049167327518e-05,
+      "loss": 1.4775,
+      "step": 5380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6666260957717896,
+      "learning_rate": 4.858022733280466e-05,
+      "loss": 1.3789,
+      "step": 5381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2422065734863281,
+      "learning_rate": 4.8579962992334126e-05,
+      "loss": 1.2948,
+      "step": 5382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.775316596031189,
+      "learning_rate": 4.8579698651863606e-05,
+      "loss": 1.4563,
+      "step": 5383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0681393146514893,
+      "learning_rate": 4.857943431139307e-05,
+      "loss": 1.5567,
+      "step": 5384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3647867441177368,
+      "learning_rate": 4.857916997092255e-05,
+      "loss": 1.3881,
+      "step": 5385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.133657932281494,
+      "learning_rate": 4.857890563045203e-05,
+      "loss": 1.4947,
+      "step": 5386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4003082513809204,
+      "learning_rate": 4.8578641289981494e-05,
+      "loss": 1.244,
+      "step": 5387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9796533584594727,
+      "learning_rate": 4.8578376949510974e-05,
+      "loss": 1.4018,
+      "step": 5388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0807647705078125,
+      "learning_rate": 4.857811260904044e-05,
+      "loss": 1.3879,
+      "step": 5389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5984512567520142,
+      "learning_rate": 4.857784826856992e-05,
+      "loss": 1.3209,
+      "step": 5390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4201927185058594,
+      "learning_rate": 4.8577583928099395e-05,
+      "loss": 1.4328,
+      "step": 5391
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.723170280456543,
+      "learning_rate": 4.857731958762887e-05,
+      "loss": 2.1977,
+      "step": 5392
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.481226682662964,
+      "learning_rate": 4.857705524715834e-05,
+      "loss": 1.4389,
+      "step": 5393
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7002568244934082,
+      "learning_rate": 4.8576790906687816e-05,
+      "loss": 1.4717,
+      "step": 5394
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7188172340393066,
+      "learning_rate": 4.857652656621729e-05,
+      "loss": 1.2621,
+      "step": 5395
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1718814373016357,
+      "learning_rate": 4.857626222574676e-05,
+      "loss": 1.4274,
+      "step": 5396
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6237578392028809,
+      "learning_rate": 4.8575997885276236e-05,
+      "loss": 1.3281,
+      "step": 5397
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9839966297149658,
+      "learning_rate": 4.857573354480571e-05,
+      "loss": 1.4654,
+      "step": 5398
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3958542346954346,
+      "learning_rate": 4.8575469204335184e-05,
+      "loss": 1.3815,
+      "step": 5399
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9210374355316162,
+      "learning_rate": 4.857520486386466e-05,
+      "loss": 1.402,
+      "step": 5400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4825485944747925,
+      "learning_rate": 4.857494052339414e-05,
+      "loss": 1.3288,
+      "step": 5401
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.001830816268921,
+      "learning_rate": 4.8574676182923604e-05,
+      "loss": 1.5078,
+      "step": 5402
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.278801441192627,
+      "learning_rate": 4.8574411842453085e-05,
+      "loss": 1.4199,
+      "step": 5403
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1077239513397217,
+      "learning_rate": 4.857414750198255e-05,
+      "loss": 1.4504,
+      "step": 5404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7069368362426758,
+      "learning_rate": 4.857388316151203e-05,
+      "loss": 1.3563,
+      "step": 5405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5458779335021973,
+      "learning_rate": 4.8573618821041505e-05,
+      "loss": 1.4111,
+      "step": 5406
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.096001625061035,
+      "learning_rate": 4.857335448057098e-05,
+      "loss": 1.6364,
+      "step": 5407
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1842386722564697,
+      "learning_rate": 4.857309014010045e-05,
+      "loss": 1.5358,
+      "step": 5408
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1496872901916504,
+      "learning_rate": 4.8572825799629926e-05,
+      "loss": 1.5009,
+      "step": 5409
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.515612006187439,
+      "learning_rate": 4.85725614591594e-05,
+      "loss": 1.3519,
+      "step": 5410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4017412662506104,
+      "learning_rate": 4.857229711868887e-05,
+      "loss": 1.4456,
+      "step": 5411
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.532381534576416,
+      "learning_rate": 4.857203277821835e-05,
+      "loss": 1.2873,
+      "step": 5412
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4654505252838135,
+      "learning_rate": 4.857176843774782e-05,
+      "loss": 1.4545,
+      "step": 5413
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.427716016769409,
+      "learning_rate": 4.8571504097277294e-05,
+      "loss": 1.4328,
+      "step": 5414
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.599678635597229,
+      "learning_rate": 4.857123975680677e-05,
+      "loss": 1.517,
+      "step": 5415
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9947052001953125,
+      "learning_rate": 4.857097541633625e-05,
+      "loss": 1.6087,
+      "step": 5416
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.097288131713867,
+      "learning_rate": 4.8570711075865715e-05,
+      "loss": 1.7437,
+      "step": 5417
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1213622093200684,
+      "learning_rate": 4.8570446735395195e-05,
+      "loss": 1.4778,
+      "step": 5418
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7584019899368286,
+      "learning_rate": 4.857018239492466e-05,
+      "loss": 1.4817,
+      "step": 5419
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6952251195907593,
+      "learning_rate": 4.856991805445414e-05,
+      "loss": 1.3974,
+      "step": 5420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8111610412597656,
+      "learning_rate": 4.8569653713983616e-05,
+      "loss": 1.3835,
+      "step": 5421
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9662145376205444,
+      "learning_rate": 4.856938937351308e-05,
+      "loss": 1.5112,
+      "step": 5422
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7712078094482422,
+      "learning_rate": 4.856912503304256e-05,
+      "loss": 1.4571,
+      "step": 5423
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6687495708465576,
+      "learning_rate": 4.856886069257203e-05,
+      "loss": 1.4852,
+      "step": 5424
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9857910871505737,
+      "learning_rate": 4.856859635210151e-05,
+      "loss": 1.3049,
+      "step": 5425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5442543029785156,
+      "learning_rate": 4.8568332011630983e-05,
+      "loss": 1.3502,
+      "step": 5426
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8275693655014038,
+      "learning_rate": 4.856806767116046e-05,
+      "loss": 1.403,
+      "step": 5427
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3765268325805664,
+      "learning_rate": 4.856780333068993e-05,
+      "loss": 1.302,
+      "step": 5428
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.197725296020508,
+      "learning_rate": 4.8567538990219404e-05,
+      "loss": 1.494,
+      "step": 5429
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.605819582939148,
+      "learning_rate": 4.856727464974888e-05,
+      "loss": 1.3051,
+      "step": 5430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8933383226394653,
+      "learning_rate": 4.856701030927835e-05,
+      "loss": 1.4934,
+      "step": 5431
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.50327467918396,
+      "learning_rate": 4.8566745968807825e-05,
+      "loss": 1.2849,
+      "step": 5432
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6833304166793823,
+      "learning_rate": 4.85664816283373e-05,
+      "loss": 1.441,
+      "step": 5433
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5800350904464722,
+      "learning_rate": 4.856621728786677e-05,
+      "loss": 1.3923,
+      "step": 5434
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0711538791656494,
+      "learning_rate": 4.8565952947396246e-05,
+      "loss": 1.3833,
+      "step": 5435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8902817964553833,
+      "learning_rate": 4.8565688606925726e-05,
+      "loss": 1.4193,
+      "step": 5436
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0076985359191895,
+      "learning_rate": 4.856542426645519e-05,
+      "loss": 1.453,
+      "step": 5437
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9239187240600586,
+      "learning_rate": 4.856515992598467e-05,
+      "loss": 1.393,
+      "step": 5438
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.962661623954773,
+      "learning_rate": 4.856489558551414e-05,
+      "loss": 1.3575,
+      "step": 5439
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9703463315963745,
+      "learning_rate": 4.856463124504362e-05,
+      "loss": 1.2677,
+      "step": 5440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4875929355621338,
+      "learning_rate": 4.8564366904573094e-05,
+      "loss": 1.3153,
+      "step": 5441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2919867038726807,
+      "learning_rate": 4.856410256410257e-05,
+      "loss": 2.4749,
+      "step": 5442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7564570903778076,
+      "learning_rate": 4.856383822363204e-05,
+      "loss": 1.4197,
+      "step": 5443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0031285285949707,
+      "learning_rate": 4.8563573883161515e-05,
+      "loss": 1.4487,
+      "step": 5444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.143563747406006,
+      "learning_rate": 4.856330954269099e-05,
+      "loss": 1.5634,
+      "step": 5445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8769463300704956,
+      "learning_rate": 4.856304520222046e-05,
+      "loss": 1.5737,
+      "step": 5446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6164716482162476,
+      "learning_rate": 4.8562780861749935e-05,
+      "loss": 1.3071,
+      "step": 5447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5084240436553955,
+      "learning_rate": 4.856251652127941e-05,
+      "loss": 1.9937,
+      "step": 5448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7743446826934814,
+      "learning_rate": 4.856225218080888e-05,
+      "loss": 1.2751,
+      "step": 5449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0147294998168945,
+      "learning_rate": 4.8561987840338356e-05,
+      "loss": 1.5426,
+      "step": 5450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.755784511566162,
+      "learning_rate": 4.8561723499867836e-05,
+      "loss": 1.4557,
+      "step": 5451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7695152759552002,
+      "learning_rate": 4.85614591593973e-05,
+      "loss": 1.3999,
+      "step": 5452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4034652709960938,
+      "learning_rate": 4.8561194818926783e-05,
+      "loss": 1.5241,
+      "step": 5453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4544286727905273,
+      "learning_rate": 4.856093047845625e-05,
+      "loss": 1.2637,
+      "step": 5454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9781911373138428,
+      "learning_rate": 4.856066613798573e-05,
+      "loss": 1.5029,
+      "step": 5455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4472200870513916,
+      "learning_rate": 4.8560401797515204e-05,
+      "loss": 1.4729,
+      "step": 5456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5427017211914062,
+      "learning_rate": 4.856013745704467e-05,
+      "loss": 1.4198,
+      "step": 5457
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.607986330986023,
+      "learning_rate": 4.855987311657415e-05,
+      "loss": 1.4893,
+      "step": 5458
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4238497018814087,
+      "learning_rate": 4.855960877610362e-05,
+      "loss": 1.3676,
+      "step": 5459
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.792710542678833,
+      "learning_rate": 4.85593444356331e-05,
+      "loss": 1.6107,
+      "step": 5460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8600834608078003,
+      "learning_rate": 4.855908009516257e-05,
+      "loss": 1.4209,
+      "step": 5461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.679725170135498,
+      "learning_rate": 4.8558815754692046e-05,
+      "loss": 1.8504,
+      "step": 5462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0033371448516846,
+      "learning_rate": 4.855855141422152e-05,
+      "loss": 1.5131,
+      "step": 5463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6065435409545898,
+      "learning_rate": 4.855828707375099e-05,
+      "loss": 2.4564,
+      "step": 5464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7562092542648315,
+      "learning_rate": 4.8558022733280466e-05,
+      "loss": 1.4361,
+      "step": 5465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.002562999725342,
+      "learning_rate": 4.855775839280995e-05,
+      "loss": 1.3803,
+      "step": 5466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6882407665252686,
+      "learning_rate": 4.8557494052339413e-05,
+      "loss": 1.2786,
+      "step": 5467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3237122297286987,
+      "learning_rate": 4.855722971186889e-05,
+      "loss": 1.39,
+      "step": 5468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8404408693313599,
+      "learning_rate": 4.855696537139836e-05,
+      "loss": 1.4876,
+      "step": 5469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.481868267059326,
+      "learning_rate": 4.8556701030927834e-05,
+      "loss": 1.6286,
+      "step": 5470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.868533968925476,
+      "learning_rate": 4.8556436690457314e-05,
+      "loss": 1.3387,
+      "step": 5471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6707919836044312,
+      "learning_rate": 4.855617234998678e-05,
+      "loss": 1.3361,
+      "step": 5472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4506070613861084,
+      "learning_rate": 4.855590800951626e-05,
+      "loss": 1.3762,
+      "step": 5473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4654219150543213,
+      "learning_rate": 4.855564366904573e-05,
+      "loss": 1.57,
+      "step": 5474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9134546518325806,
+      "learning_rate": 4.855537932857521e-05,
+      "loss": 1.3879,
+      "step": 5475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7851284742355347,
+      "learning_rate": 4.855511498810468e-05,
+      "loss": 1.3845,
+      "step": 5476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.364372730255127,
+      "learning_rate": 4.8554850647634156e-05,
+      "loss": 1.685,
+      "step": 5477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.581644058227539,
+      "learning_rate": 4.855458630716363e-05,
+      "loss": 1.4833,
+      "step": 5478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.706502914428711,
+      "learning_rate": 4.85543219666931e-05,
+      "loss": 1.3761,
+      "step": 5479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.932352900505066,
+      "learning_rate": 4.855405762622258e-05,
+      "loss": 1.3194,
+      "step": 5480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8561224937438965,
+      "learning_rate": 4.855379328575205e-05,
+      "loss": 1.3701,
+      "step": 5481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7388864755630493,
+      "learning_rate": 4.8553528945281524e-05,
+      "loss": 1.4746,
+      "step": 5482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7615090608596802,
+      "learning_rate": 4.8553264604811e-05,
+      "loss": 2.0206,
+      "step": 5483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.841283917427063,
+      "learning_rate": 4.855300026434047e-05,
+      "loss": 1.4231,
+      "step": 5484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7538838386535645,
+      "learning_rate": 4.8552735923869944e-05,
+      "loss": 1.3822,
+      "step": 5485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.97415292263031,
+      "learning_rate": 4.8552471583399425e-05,
+      "loss": 1.3911,
+      "step": 5486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6137797832489014,
+      "learning_rate": 4.855220724292889e-05,
+      "loss": 1.4191,
+      "step": 5487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.796779990196228,
+      "learning_rate": 4.855194290245837e-05,
+      "loss": 1.4354,
+      "step": 5488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6877589225769043,
+      "learning_rate": 4.855167856198784e-05,
+      "loss": 1.4848,
+      "step": 5489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3773009777069092,
+      "learning_rate": 4.855141422151732e-05,
+      "loss": 1.4127,
+      "step": 5490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4307469129562378,
+      "learning_rate": 4.855114988104679e-05,
+      "loss": 1.238,
+      "step": 5491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9288136959075928,
+      "learning_rate": 4.8550885540576266e-05,
+      "loss": 1.4932,
+      "step": 5492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9191094636917114,
+      "learning_rate": 4.855062120010574e-05,
+      "loss": 1.3588,
+      "step": 5493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.990037441253662,
+      "learning_rate": 4.855035685963521e-05,
+      "loss": 1.3467,
+      "step": 5494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3728761672973633,
+      "learning_rate": 4.855009251916469e-05,
+      "loss": 1.6586,
+      "step": 5495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7948943376541138,
+      "learning_rate": 4.854982817869416e-05,
+      "loss": 1.4738,
+      "step": 5496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2721431255340576,
+      "learning_rate": 4.8549563838223634e-05,
+      "loss": 1.3816,
+      "step": 5497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1535980701446533,
+      "learning_rate": 4.854929949775311e-05,
+      "loss": 1.3754,
+      "step": 5498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7716338634490967,
+      "learning_rate": 4.854903515728258e-05,
+      "loss": 1.4027,
+      "step": 5499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6933526992797852,
+      "learning_rate": 4.8548770816812055e-05,
+      "loss": 1.2113,
+      "step": 5500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8662458658218384,
+      "learning_rate": 4.8548506476341535e-05,
+      "loss": 1.4092,
+      "step": 5501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.282184600830078,
+      "learning_rate": 4.8548242135871e-05,
+      "loss": 1.6021,
+      "step": 5502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7260148525238037,
+      "learning_rate": 4.8547977795400476e-05,
+      "loss": 1.3979,
+      "step": 5503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.590517282485962,
+      "learning_rate": 4.854771345492995e-05,
+      "loss": 1.3326,
+      "step": 5504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1984341144561768,
+      "learning_rate": 4.854744911445942e-05,
+      "loss": 1.3493,
+      "step": 5505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.132878065109253,
+      "learning_rate": 4.85471847739889e-05,
+      "loss": 1.5061,
+      "step": 5506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.118834972381592,
+      "learning_rate": 4.854692043351837e-05,
+      "loss": 1.8214,
+      "step": 5507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9323257207870483,
+      "learning_rate": 4.854665609304785e-05,
+      "loss": 1.3817,
+      "step": 5508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.647117018699646,
+      "learning_rate": 4.854639175257732e-05,
+      "loss": 1.3289,
+      "step": 5509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9087610244750977,
+      "learning_rate": 4.85461274121068e-05,
+      "loss": 1.3293,
+      "step": 5510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.405695676803589,
+      "learning_rate": 4.854586307163627e-05,
+      "loss": 1.4488,
+      "step": 5511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9209272861480713,
+      "learning_rate": 4.8545598731165744e-05,
+      "loss": 1.5211,
+      "step": 5512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3593966960906982,
+      "learning_rate": 4.854533439069522e-05,
+      "loss": 1.3051,
+      "step": 5513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4113267660140991,
+      "learning_rate": 4.854507005022469e-05,
+      "loss": 1.3079,
+      "step": 5514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6924546957015991,
+      "learning_rate": 4.8544805709754165e-05,
+      "loss": 1.4809,
+      "step": 5515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.047461748123169,
+      "learning_rate": 4.854454136928364e-05,
+      "loss": 1.3892,
+      "step": 5516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6433720588684082,
+      "learning_rate": 4.854427702881311e-05,
+      "loss": 1.3773,
+      "step": 5517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.646470069885254,
+      "learning_rate": 4.8544012688342586e-05,
+      "loss": 1.4867,
+      "step": 5518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7802188396453857,
+      "learning_rate": 4.854374834787206e-05,
+      "loss": 1.4191,
+      "step": 5519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.711496114730835,
+      "learning_rate": 4.854348400740153e-05,
+      "loss": 1.3153,
+      "step": 5520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8853789567947388,
+      "learning_rate": 4.854321966693101e-05,
+      "loss": 1.3038,
+      "step": 5521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.55180025100708,
+      "learning_rate": 4.854295532646048e-05,
+      "loss": 1.3724,
+      "step": 5522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8237838745117188,
+      "learning_rate": 4.854269098598996e-05,
+      "loss": 1.3177,
+      "step": 5523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6052742004394531,
+      "learning_rate": 4.854242664551943e-05,
+      "loss": 1.2775,
+      "step": 5524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.768040418624878,
+      "learning_rate": 4.854216230504891e-05,
+      "loss": 1.4528,
+      "step": 5525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8174121379852295,
+      "learning_rate": 4.854189796457838e-05,
+      "loss": 1.5871,
+      "step": 5526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.224648952484131,
+      "learning_rate": 4.8541633624107855e-05,
+      "loss": 1.5503,
+      "step": 5527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0558290481567383,
+      "learning_rate": 4.854136928363733e-05,
+      "loss": 1.7432,
+      "step": 5528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.924932837486267,
+      "learning_rate": 4.85411049431668e-05,
+      "loss": 1.3849,
+      "step": 5529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8089005947113037,
+      "learning_rate": 4.8540840602696276e-05,
+      "loss": 1.4514,
+      "step": 5530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7368528842926025,
+      "learning_rate": 4.854057626222575e-05,
+      "loss": 1.5274,
+      "step": 5531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.025678873062134,
+      "learning_rate": 4.854031192175522e-05,
+      "loss": 1.3474,
+      "step": 5532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3322107791900635,
+      "learning_rate": 4.8540047581284696e-05,
+      "loss": 1.4614,
+      "step": 5533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6491239070892334,
+      "learning_rate": 4.853978324081417e-05,
+      "loss": 2.4771,
+      "step": 5534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6404798030853271,
+      "learning_rate": 4.853951890034364e-05,
+      "loss": 1.504,
+      "step": 5535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6571435928344727,
+      "learning_rate": 4.8539254559873124e-05,
+      "loss": 1.3645,
+      "step": 5536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8500028848648071,
+      "learning_rate": 4.853899021940259e-05,
+      "loss": 1.4954,
+      "step": 5537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7100692987442017,
+      "learning_rate": 4.853872587893207e-05,
+      "loss": 1.9862,
+      "step": 5538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6232824325561523,
+      "learning_rate": 4.853846153846154e-05,
+      "loss": 1.3837,
+      "step": 5539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7208737134933472,
+      "learning_rate": 4.853819719799101e-05,
+      "loss": 1.4544,
+      "step": 5540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6277434825897217,
+      "learning_rate": 4.853793285752049e-05,
+      "loss": 1.5906,
+      "step": 5541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.479299783706665,
+      "learning_rate": 4.853766851704996e-05,
+      "loss": 1.2507,
+      "step": 5542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3875458240509033,
+      "learning_rate": 4.853740417657944e-05,
+      "loss": 1.4723,
+      "step": 5543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0672409534454346,
+      "learning_rate": 4.8537139836108906e-05,
+      "loss": 1.3819,
+      "step": 5544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.784358263015747,
+      "learning_rate": 4.8536875495638386e-05,
+      "loss": 1.4551,
+      "step": 5545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.188706636428833,
+      "learning_rate": 4.853661115516786e-05,
+      "loss": 1.4569,
+      "step": 5546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4411733150482178,
+      "learning_rate": 4.853634681469733e-05,
+      "loss": 1.3005,
+      "step": 5547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.487434983253479,
+      "learning_rate": 4.8536082474226807e-05,
+      "loss": 1.3962,
+      "step": 5548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5529378652572632,
+      "learning_rate": 4.853581813375628e-05,
+      "loss": 1.2757,
+      "step": 5549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6535673141479492,
+      "learning_rate": 4.8535553793285754e-05,
+      "loss": 2.5114,
+      "step": 5550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9615812301635742,
+      "learning_rate": 4.853528945281523e-05,
+      "loss": 1.283,
+      "step": 5551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5003620386123657,
+      "learning_rate": 4.85350251123447e-05,
+      "loss": 1.3139,
+      "step": 5552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.152980327606201,
+      "learning_rate": 4.8534760771874174e-05,
+      "loss": 1.4332,
+      "step": 5553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.718792200088501,
+      "learning_rate": 4.853449643140365e-05,
+      "loss": 1.8713,
+      "step": 5554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7670234441757202,
+      "learning_rate": 4.853423209093312e-05,
+      "loss": 1.3844,
+      "step": 5555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.633839726448059,
+      "learning_rate": 4.85339677504626e-05,
+      "loss": 1.3533,
+      "step": 5556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7035527229309082,
+      "learning_rate": 4.853370340999207e-05,
+      "loss": 1.4275,
+      "step": 5557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6703426837921143,
+      "learning_rate": 4.853343906952155e-05,
+      "loss": 1.3755,
+      "step": 5558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.178023815155029,
+      "learning_rate": 4.8533174729051016e-05,
+      "loss": 1.4004,
+      "step": 5559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.457919955253601,
+      "learning_rate": 4.8532910388580496e-05,
+      "loss": 1.2849,
+      "step": 5560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6101223230361938,
+      "learning_rate": 4.853264604810997e-05,
+      "loss": 1.4376,
+      "step": 5561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4894402027130127,
+      "learning_rate": 4.853238170763944e-05,
+      "loss": 1.5517,
+      "step": 5562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.707651138305664,
+      "learning_rate": 4.853211736716892e-05,
+      "loss": 1.3766,
+      "step": 5563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.815173864364624,
+      "learning_rate": 4.853185302669839e-05,
+      "loss": 1.3807,
+      "step": 5564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7392348051071167,
+      "learning_rate": 4.8531588686227864e-05,
+      "loss": 1.4387,
+      "step": 5565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7326890230178833,
+      "learning_rate": 4.853132434575734e-05,
+      "loss": 1.4073,
+      "step": 5566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7070538997650146,
+      "learning_rate": 4.853106000528681e-05,
+      "loss": 1.3564,
+      "step": 5567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1210575103759766,
+      "learning_rate": 4.8530795664816285e-05,
+      "loss": 1.4844,
+      "step": 5568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7462027072906494,
+      "learning_rate": 4.853053132434576e-05,
+      "loss": 1.3893,
+      "step": 5569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8725414276123047,
+      "learning_rate": 4.853026698387523e-05,
+      "loss": 1.4631,
+      "step": 5570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.013777017593384,
+      "learning_rate": 4.853000264340471e-05,
+      "loss": 1.378,
+      "step": 5571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.675222396850586,
+      "learning_rate": 4.852973830293418e-05,
+      "loss": 1.282,
+      "step": 5572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0830159187316895,
+      "learning_rate": 4.852947396246366e-05,
+      "loss": 1.4215,
+      "step": 5573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8926875591278076,
+      "learning_rate": 4.8529209621993126e-05,
+      "loss": 1.4932,
+      "step": 5574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0769481658935547,
+      "learning_rate": 4.85289452815226e-05,
+      "loss": 1.5999,
+      "step": 5575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6673941612243652,
+      "learning_rate": 4.852868094105208e-05,
+      "loss": 1.3855,
+      "step": 5576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.482321858406067,
+      "learning_rate": 4.852841660058155e-05,
+      "loss": 1.3077,
+      "step": 5577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.717428207397461,
+      "learning_rate": 4.852815226011103e-05,
+      "loss": 1.2855,
+      "step": 5578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.624631404876709,
+      "learning_rate": 4.8527887919640494e-05,
+      "loss": 1.3911,
+      "step": 5579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5454002618789673,
+      "learning_rate": 4.8527623579169974e-05,
+      "loss": 1.3867,
+      "step": 5580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.594256043434143,
+      "learning_rate": 4.852735923869945e-05,
+      "loss": 1.4928,
+      "step": 5581
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.07094144821167,
+      "learning_rate": 4.852709489822892e-05,
+      "loss": 1.5085,
+      "step": 5582
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5861958265304565,
+      "learning_rate": 4.8526830557758395e-05,
+      "loss": 1.3933,
+      "step": 5583
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.746362328529358,
+      "learning_rate": 4.852656621728787e-05,
+      "loss": 1.3675,
+      "step": 5584
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6542690992355347,
+      "learning_rate": 4.852630187681734e-05,
+      "loss": 1.3161,
+      "step": 5585
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.815978765487671,
+      "learning_rate": 4.8526037536346816e-05,
+      "loss": 1.4721,
+      "step": 5586
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.176165819168091,
+      "learning_rate": 4.852577319587629e-05,
+      "loss": 1.5925,
+      "step": 5587
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.872877836227417,
+      "learning_rate": 4.852550885540576e-05,
+      "loss": 1.4488,
+      "step": 5588
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9124678373336792,
+      "learning_rate": 4.8525244514935237e-05,
+      "loss": 1.18,
+      "step": 5589
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.888185977935791,
+      "learning_rate": 4.852498017446471e-05,
+      "loss": 1.5874,
+      "step": 5590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7327728271484375,
+      "learning_rate": 4.852471583399419e-05,
+      "loss": 1.2567,
+      "step": 5591
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.815131664276123,
+      "learning_rate": 4.852445149352366e-05,
+      "loss": 1.4346,
+      "step": 5592
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.502476215362549,
+      "learning_rate": 4.852418715305314e-05,
+      "loss": 1.4829,
+      "step": 5593
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4852046966552734,
+      "learning_rate": 4.8523922812582604e-05,
+      "loss": 1.5828,
+      "step": 5594
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7562774419784546,
+      "learning_rate": 4.8523658472112085e-05,
+      "loss": 1.3492,
+      "step": 5595
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7319722175598145,
+      "learning_rate": 4.852339413164156e-05,
+      "loss": 1.313,
+      "step": 5596
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1992228031158447,
+      "learning_rate": 4.852312979117103e-05,
+      "loss": 1.3824,
+      "step": 5597
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3214187622070312,
+      "learning_rate": 4.8522865450700505e-05,
+      "loss": 1.2878,
+      "step": 5598
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7949273586273193,
+      "learning_rate": 4.852260111022998e-05,
+      "loss": 1.6371,
+      "step": 5599
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.127513885498047,
+      "learning_rate": 4.852233676975945e-05,
+      "loss": 1.5624,
+      "step": 5600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.773103952407837,
+      "learning_rate": 4.8522072429288926e-05,
+      "loss": 1.2829,
+      "step": 5601
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5450818538665771,
+      "learning_rate": 4.85218080888184e-05,
+      "loss": 1.442,
+      "step": 5602
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7777564525604248,
+      "learning_rate": 4.852154374834787e-05,
+      "loss": 1.3498,
+      "step": 5603
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.629582166671753,
+      "learning_rate": 4.852127940787735e-05,
+      "loss": 1.374,
+      "step": 5604
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7795405387878418,
+      "learning_rate": 4.852101506740682e-05,
+      "loss": 1.4488,
+      "step": 5605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.117863178253174,
+      "learning_rate": 4.85207507269363e-05,
+      "loss": 1.3515,
+      "step": 5606
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.405531883239746,
+      "learning_rate": 4.852048638646577e-05,
+      "loss": 1.2879,
+      "step": 5607
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8485023975372314,
+      "learning_rate": 4.852022204599525e-05,
+      "loss": 1.3803,
+      "step": 5608
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8494367599487305,
+      "learning_rate": 4.8519957705524715e-05,
+      "loss": 1.358,
+      "step": 5609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3415112495422363,
+      "learning_rate": 4.8519693365054195e-05,
+      "loss": 1.257,
+      "step": 5610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.946995735168457,
+      "learning_rate": 4.851942902458367e-05,
+      "loss": 1.606,
+      "step": 5611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.845611810684204,
+      "learning_rate": 4.8519164684113135e-05,
+      "loss": 1.3951,
+      "step": 5612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5584690570831299,
+      "learning_rate": 4.8518900343642616e-05,
+      "loss": 1.3375,
+      "step": 5613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.593670129776001,
+      "learning_rate": 4.851863600317208e-05,
+      "loss": 1.4121,
+      "step": 5614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.000523805618286,
+      "learning_rate": 4.851837166270156e-05,
+      "loss": 1.3703,
+      "step": 5615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5490490198135376,
+      "learning_rate": 4.8518107322231036e-05,
+      "loss": 1.4832,
+      "step": 5616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3070751428604126,
+      "learning_rate": 4.851784298176051e-05,
+      "loss": 1.6445,
+      "step": 5617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8896546363830566,
+      "learning_rate": 4.8517578641289984e-05,
+      "loss": 1.5044,
+      "step": 5618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2403435707092285,
+      "learning_rate": 4.851731430081946e-05,
+      "loss": 1.6013,
+      "step": 5619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7327688932418823,
+      "learning_rate": 4.851704996034893e-05,
+      "loss": 1.4657,
+      "step": 5620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6514077186584473,
+      "learning_rate": 4.8516785619878404e-05,
+      "loss": 1.423,
+      "step": 5621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2185134887695312,
+      "learning_rate": 4.851652127940788e-05,
+      "loss": 1.4996,
+      "step": 5622
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9518375396728516,
+      "learning_rate": 4.851625693893735e-05,
+      "loss": 1.5207,
+      "step": 5623
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4015408754348755,
+      "learning_rate": 4.8515992598466825e-05,
+      "loss": 1.2774,
+      "step": 5624
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.010744094848633,
+      "learning_rate": 4.85157282579963e-05,
+      "loss": 1.4957,
+      "step": 5625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9012272357940674,
+      "learning_rate": 4.851546391752578e-05,
+      "loss": 1.3028,
+      "step": 5626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7878352403640747,
+      "learning_rate": 4.8515199577055246e-05,
+      "loss": 1.4455,
+      "step": 5627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5582505464553833,
+      "learning_rate": 4.8514935236584726e-05,
+      "loss": 1.409,
+      "step": 5628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6012909412384033,
+      "learning_rate": 4.851467089611419e-05,
+      "loss": 1.4564,
+      "step": 5629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5011142492294312,
+      "learning_rate": 4.851440655564367e-05,
+      "loss": 1.318,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.569699764251709,
+      "learning_rate": 4.851414221517315e-05,
+      "loss": 1.3051,
+      "step": 5631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7745493650436401,
+      "learning_rate": 4.851387787470262e-05,
+      "loss": 1.3931,
+      "step": 5632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0661113262176514,
+      "learning_rate": 4.8513613534232094e-05,
+      "loss": 1.4899,
+      "step": 5633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5360990762710571,
+      "learning_rate": 4.851334919376157e-05,
+      "loss": 1.3323,
+      "step": 5634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.501349687576294,
+      "learning_rate": 4.851308485329104e-05,
+      "loss": 1.5147,
+      "step": 5635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.561835765838623,
+      "learning_rate": 4.8512820512820515e-05,
+      "loss": 1.4288,
+      "step": 5636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9555914402008057,
+      "learning_rate": 4.851255617234999e-05,
+      "loss": 1.2836,
+      "step": 5637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6965222358703613,
+      "learning_rate": 4.851229183187946e-05,
+      "loss": 1.3427,
+      "step": 5638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.684216856956482,
+      "learning_rate": 4.8512027491408935e-05,
+      "loss": 1.3851,
+      "step": 5639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8990148305892944,
+      "learning_rate": 4.851176315093841e-05,
+      "loss": 1.431,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.79413640499115,
+      "learning_rate": 4.851149881046789e-05,
+      "loss": 1.4207,
+      "step": 5641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6171014308929443,
+      "learning_rate": 4.8511234469997356e-05,
+      "loss": 1.5198,
+      "step": 5642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3041696548461914,
+      "learning_rate": 4.8510970129526836e-05,
+      "loss": 1.2503,
+      "step": 5643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9277263879776,
+      "learning_rate": 4.85107057890563e-05,
+      "loss": 1.494,
+      "step": 5644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6833300590515137,
+      "learning_rate": 4.8510441448585784e-05,
+      "loss": 1.2471,
+      "step": 5645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2973089218139648,
+      "learning_rate": 4.851017710811526e-05,
+      "loss": 1.3052,
+      "step": 5646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.498762845993042,
+      "learning_rate": 4.850991276764473e-05,
+      "loss": 1.3564,
+      "step": 5647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.777597427368164,
+      "learning_rate": 4.8509648427174204e-05,
+      "loss": 1.3333,
+      "step": 5648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.199558973312378,
+      "learning_rate": 4.850938408670367e-05,
+      "loss": 1.3802,
+      "step": 5649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9792730808258057,
+      "learning_rate": 4.850911974623315e-05,
+      "loss": 1.6708,
+      "step": 5650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.157564640045166,
+      "learning_rate": 4.8508855405762625e-05,
+      "loss": 1.3374,
+      "step": 5651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.258502244949341,
+      "learning_rate": 4.85085910652921e-05,
+      "loss": 1.357,
+      "step": 5652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.710439682006836,
+      "learning_rate": 4.850832672482157e-05,
+      "loss": 1.8979,
+      "step": 5653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8636431694030762,
+      "learning_rate": 4.8508062384351046e-05,
+      "loss": 1.9042,
+      "step": 5654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1426122188568115,
+      "learning_rate": 4.850779804388052e-05,
+      "loss": 1.4692,
+      "step": 5655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7270286083221436,
+      "learning_rate": 4.850753370341e-05,
+      "loss": 1.2803,
+      "step": 5656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.189932346343994,
+      "learning_rate": 4.8507269362939466e-05,
+      "loss": 1.7791,
+      "step": 5657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.329596757888794,
+      "learning_rate": 4.850700502246894e-05,
+      "loss": 1.5518,
+      "step": 5658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1219918727874756,
+      "learning_rate": 4.8506740681998414e-05,
+      "loss": 1.2709,
+      "step": 5659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9105043411254883,
+      "learning_rate": 4.850647634152789e-05,
+      "loss": 1.49,
+      "step": 5660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.869571566581726,
+      "learning_rate": 4.850621200105737e-05,
+      "loss": 1.5454,
+      "step": 5661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9629496335983276,
+      "learning_rate": 4.8505947660586834e-05,
+      "loss": 1.4665,
+      "step": 5662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6951693296432495,
+      "learning_rate": 4.8505683320116315e-05,
+      "loss": 1.4621,
+      "step": 5663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9369233846664429,
+      "learning_rate": 4.850541897964578e-05,
+      "loss": 1.4215,
+      "step": 5664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.875957727432251,
+      "learning_rate": 4.850515463917526e-05,
+      "loss": 1.4192,
+      "step": 5665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6046847105026245,
+      "learning_rate": 4.8504890298704735e-05,
+      "loss": 1.2126,
+      "step": 5666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.008009910583496,
+      "learning_rate": 4.850462595823421e-05,
+      "loss": 1.4988,
+      "step": 5667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8434683084487915,
+      "learning_rate": 4.850436161776368e-05,
+      "loss": 1.3834,
+      "step": 5668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6923701763153076,
+      "learning_rate": 4.8504097277293156e-05,
+      "loss": 1.3717,
+      "step": 5669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5864098072052,
+      "learning_rate": 4.850383293682263e-05,
+      "loss": 1.5736,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8864799737930298,
+      "learning_rate": 4.85035685963521e-05,
+      "loss": 1.2681,
+      "step": 5671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6854859590530396,
+      "learning_rate": 4.850330425588158e-05,
+      "loss": 1.4562,
+      "step": 5672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5327612161636353,
+      "learning_rate": 4.850303991541105e-05,
+      "loss": 1.3809,
+      "step": 5673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0624430179595947,
+      "learning_rate": 4.8502775574940524e-05,
+      "loss": 1.516,
+      "step": 5674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7987037897109985,
+      "learning_rate": 4.850251123447e-05,
+      "loss": 1.4382,
+      "step": 5675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4363586902618408,
+      "learning_rate": 4.850224689399948e-05,
+      "loss": 1.2911,
+      "step": 5676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8822935819625854,
+      "learning_rate": 4.8501982553528945e-05,
+      "loss": 1.3782,
+      "step": 5677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4655603170394897,
+      "learning_rate": 4.8501718213058425e-05,
+      "loss": 1.4351,
+      "step": 5678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.705259919166565,
+      "learning_rate": 4.850145387258789e-05,
+      "loss": 1.5684,
+      "step": 5679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.780163288116455,
+      "learning_rate": 4.850118953211737e-05,
+      "loss": 1.3888,
+      "step": 5680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4829403162002563,
+      "learning_rate": 4.8500925191646846e-05,
+      "loss": 1.4428,
+      "step": 5681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6825017929077148,
+      "learning_rate": 4.850066085117632e-05,
+      "loss": 1.3157,
+      "step": 5682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4535932540893555,
+      "learning_rate": 4.850039651070579e-05,
+      "loss": 1.3827,
+      "step": 5683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.404862880706787,
+      "learning_rate": 4.850013217023526e-05,
+      "loss": 1.6158,
+      "step": 5684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.93293297290802,
+      "learning_rate": 4.849986782976474e-05,
+      "loss": 1.6382,
+      "step": 5685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7461657524108887,
+      "learning_rate": 4.8499603489294214e-05,
+      "loss": 2.2788,
+      "step": 5686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.694612741470337,
+      "learning_rate": 4.849933914882369e-05,
+      "loss": 1.3218,
+      "step": 5687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6092798709869385,
+      "learning_rate": 4.849907480835316e-05,
+      "loss": 1.3438,
+      "step": 5688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6446211338043213,
+      "learning_rate": 4.8498810467882634e-05,
+      "loss": 1.5229,
+      "step": 5689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4326056241989136,
+      "learning_rate": 4.849854612741211e-05,
+      "loss": 1.3472,
+      "step": 5690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.554438591003418,
+      "learning_rate": 4.849828178694159e-05,
+      "loss": 1.3508,
+      "step": 5691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8544175624847412,
+      "learning_rate": 4.8498017446471055e-05,
+      "loss": 1.3076,
+      "step": 5692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1223020553588867,
+      "learning_rate": 4.849775310600053e-05,
+      "loss": 1.5133,
+      "step": 5693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6842565536499023,
+      "learning_rate": 4.849748876553e-05,
+      "loss": 2.2305,
+      "step": 5694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5435655117034912,
+      "learning_rate": 4.8497224425059476e-05,
+      "loss": 2.1234,
+      "step": 5695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2634763717651367,
+      "learning_rate": 4.8496960084588956e-05,
+      "loss": 1.3753,
+      "step": 5696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.637652039527893,
+      "learning_rate": 4.849669574411842e-05,
+      "loss": 1.3551,
+      "step": 5697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.681143879890442,
+      "learning_rate": 4.84964314036479e-05,
+      "loss": 1.3474,
+      "step": 5698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9378314018249512,
+      "learning_rate": 4.849616706317737e-05,
+      "loss": 1.4577,
+      "step": 5699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9117246866226196,
+      "learning_rate": 4.849590272270685e-05,
+      "loss": 1.4857,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4435746669769287,
+      "learning_rate": 4.8495638382236324e-05,
+      "loss": 1.3359,
+      "step": 5701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7056564092636108,
+      "learning_rate": 4.84953740417658e-05,
+      "loss": 2.2253,
+      "step": 5702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.550323247909546,
+      "learning_rate": 4.849510970129527e-05,
+      "loss": 1.2226,
+      "step": 5703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3666434288024902,
+      "learning_rate": 4.8494845360824745e-05,
+      "loss": 1.4205,
+      "step": 5704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.094661712646484,
+      "learning_rate": 4.849458102035422e-05,
+      "loss": 1.5009,
+      "step": 5705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4081779718399048,
+      "learning_rate": 4.849431667988369e-05,
+      "loss": 1.2805,
+      "step": 5706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5960938930511475,
+      "learning_rate": 4.8494052339413165e-05,
+      "loss": 1.3392,
+      "step": 5707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6896028518676758,
+      "learning_rate": 4.849378799894264e-05,
+      "loss": 1.4444,
+      "step": 5708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.405027151107788,
+      "learning_rate": 4.849352365847211e-05,
+      "loss": 1.374,
+      "step": 5709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8795257806777954,
+      "learning_rate": 4.8493259318001586e-05,
+      "loss": 1.3669,
+      "step": 5710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9489192962646484,
+      "learning_rate": 4.8492994977531066e-05,
+      "loss": 1.3518,
+      "step": 5711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5272091627120972,
+      "learning_rate": 4.849273063706053e-05,
+      "loss": 1.3048,
+      "step": 5712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5377014875411987,
+      "learning_rate": 4.8492466296590014e-05,
+      "loss": 1.3294,
+      "step": 5713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7321484088897705,
+      "learning_rate": 4.849220195611948e-05,
+      "loss": 1.4335,
+      "step": 5714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7990838289260864,
+      "learning_rate": 4.849193761564896e-05,
+      "loss": 1.5519,
+      "step": 5715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.177564859390259,
+      "learning_rate": 4.8491673275178434e-05,
+      "loss": 1.5482,
+      "step": 5716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.594992995262146,
+      "learning_rate": 4.849140893470791e-05,
+      "loss": 1.4308,
+      "step": 5717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5230594873428345,
+      "learning_rate": 4.849114459423738e-05,
+      "loss": 1.327,
+      "step": 5718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0374999046325684,
+      "learning_rate": 4.8490880253766855e-05,
+      "loss": 1.2382,
+      "step": 5719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.241842746734619,
+      "learning_rate": 4.849061591329633e-05,
+      "loss": 1.5233,
+      "step": 5720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.652883291244507,
+      "learning_rate": 4.84903515728258e-05,
+      "loss": 1.525,
+      "step": 5721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0490477085113525,
+      "learning_rate": 4.8490087232355276e-05,
+      "loss": 1.4962,
+      "step": 5722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4195536375045776,
+      "learning_rate": 4.848982289188475e-05,
+      "loss": 1.8769,
+      "step": 5723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4052023887634277,
+      "learning_rate": 4.848955855141422e-05,
+      "loss": 1.6061,
+      "step": 5724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7003490924835205,
+      "learning_rate": 4.8489294210943696e-05,
+      "loss": 1.266,
+      "step": 5725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6464738845825195,
+      "learning_rate": 4.848902987047318e-05,
+      "loss": 1.5244,
+      "step": 5726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.362065076828003,
+      "learning_rate": 4.8488765530002644e-05,
+      "loss": 1.0568,
+      "step": 5727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9955898523330688,
+      "learning_rate": 4.8488501189532124e-05,
+      "loss": 1.5628,
+      "step": 5728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7518677711486816,
+      "learning_rate": 4.848823684906159e-05,
+      "loss": 1.4319,
+      "step": 5729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1078052520751953,
+      "learning_rate": 4.8487972508591064e-05,
+      "loss": 1.4051,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9807623624801636,
+      "learning_rate": 4.8487708168120545e-05,
+      "loss": 1.382,
+      "step": 5731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4834423065185547,
+      "learning_rate": 4.848744382765001e-05,
+      "loss": 1.5536,
+      "step": 5732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7082765102386475,
+      "learning_rate": 4.848717948717949e-05,
+      "loss": 1.4591,
+      "step": 5733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.542806386947632,
+      "learning_rate": 4.848691514670896e-05,
+      "loss": 1.5046,
+      "step": 5734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.789194107055664,
+      "learning_rate": 4.848665080623844e-05,
+      "loss": 1.5696,
+      "step": 5735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.030488967895508,
+      "learning_rate": 4.848638646576791e-05,
+      "loss": 1.4123,
+      "step": 5736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8418302536010742,
+      "learning_rate": 4.8486122125297386e-05,
+      "loss": 1.4978,
+      "step": 5737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7432119846343994,
+      "learning_rate": 4.848585778482686e-05,
+      "loss": 1.426,
+      "step": 5738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7639061212539673,
+      "learning_rate": 4.848559344435633e-05,
+      "loss": 1.4338,
+      "step": 5739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6885533332824707,
+      "learning_rate": 4.848532910388581e-05,
+      "loss": 1.411,
+      "step": 5740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4934269189834595,
+      "learning_rate": 4.848506476341528e-05,
+      "loss": 1.2097,
+      "step": 5741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.408664584159851,
+      "learning_rate": 4.8484800422944754e-05,
+      "loss": 2.2043,
+      "step": 5742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7621022462844849,
+      "learning_rate": 4.848453608247423e-05,
+      "loss": 1.3883,
+      "step": 5743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.015228509902954,
+      "learning_rate": 4.84842717420037e-05,
+      "loss": 1.5305,
+      "step": 5744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.147440195083618,
+      "learning_rate": 4.8484007401533175e-05,
+      "loss": 1.2928,
+      "step": 5745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6778301000595093,
+      "learning_rate": 4.8483743061062655e-05,
+      "loss": 1.2006,
+      "step": 5746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8014601469039917,
+      "learning_rate": 4.848347872059212e-05,
+      "loss": 1.4467,
+      "step": 5747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.061211585998535,
+      "learning_rate": 4.84832143801216e-05,
+      "loss": 1.4596,
+      "step": 5748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5487089157104492,
+      "learning_rate": 4.848295003965107e-05,
+      "loss": 1.4825,
+      "step": 5749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5013267993927002,
+      "learning_rate": 4.848268569918055e-05,
+      "loss": 1.3283,
+      "step": 5750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.599318504333496,
+      "learning_rate": 4.848242135871002e-05,
+      "loss": 1.3523,
+      "step": 5751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9221441745758057,
+      "learning_rate": 4.8482157018239496e-05,
+      "loss": 1.4703,
+      "step": 5752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.970317006111145,
+      "learning_rate": 4.848189267776897e-05,
+      "loss": 1.3779,
+      "step": 5753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.967729091644287,
+      "learning_rate": 4.8481628337298443e-05,
+      "loss": 1.4425,
+      "step": 5754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6574863195419312,
+      "learning_rate": 4.848136399682792e-05,
+      "loss": 1.4419,
+      "step": 5755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7532066106796265,
+      "learning_rate": 4.848109965635739e-05,
+      "loss": 1.3587,
+      "step": 5756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4566090106964111,
+      "learning_rate": 4.8480835315886864e-05,
+      "loss": 1.28,
+      "step": 5757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.856440782546997,
+      "learning_rate": 4.848057097541634e-05,
+      "loss": 2.0229,
+      "step": 5758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6600302457809448,
+      "learning_rate": 4.848030663494581e-05,
+      "loss": 1.5741,
+      "step": 5759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4827545881271362,
+      "learning_rate": 4.8480042294475285e-05,
+      "loss": 1.4211,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.017812490463257,
+      "learning_rate": 4.8479777954004765e-05,
+      "loss": 1.4351,
+      "step": 5761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8868392705917358,
+      "learning_rate": 4.847951361353423e-05,
+      "loss": 1.348,
+      "step": 5762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.676825761795044,
+      "learning_rate": 4.847924927306371e-05,
+      "loss": 1.3947,
+      "step": 5763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.48415470123291,
+      "learning_rate": 4.847898493259318e-05,
+      "loss": 1.637,
+      "step": 5764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.10050892829895,
+      "learning_rate": 4.847872059212265e-05,
+      "loss": 1.4106,
+      "step": 5765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0672576427459717,
+      "learning_rate": 4.847845625165213e-05,
+      "loss": 1.667,
+      "step": 5766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4739902019500732,
+      "learning_rate": 4.84781919111816e-05,
+      "loss": 1.3071,
+      "step": 5767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6065267324447632,
+      "learning_rate": 4.847792757071108e-05,
+      "loss": 1.3775,
+      "step": 5768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4294825792312622,
+      "learning_rate": 4.847766323024055e-05,
+      "loss": 1.3491,
+      "step": 5769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8745280504226685,
+      "learning_rate": 4.847739888977003e-05,
+      "loss": 1.4291,
+      "step": 5770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2725672721862793,
+      "learning_rate": 4.84771345492995e-05,
+      "loss": 1.413,
+      "step": 5771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.67756187915802,
+      "learning_rate": 4.8476870208828975e-05,
+      "loss": 1.3095,
+      "step": 5772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3712518215179443,
+      "learning_rate": 4.847660586835845e-05,
+      "loss": 1.2559,
+      "step": 5773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2951135635375977,
+      "learning_rate": 4.847634152788792e-05,
+      "loss": 1.6147,
+      "step": 5774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.174143075942993,
+      "learning_rate": 4.8476077187417395e-05,
+      "loss": 1.4543,
+      "step": 5775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8605154752731323,
+      "learning_rate": 4.847581284694687e-05,
+      "loss": 1.4066,
+      "step": 5776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8067213296890259,
+      "learning_rate": 4.847554850647634e-05,
+      "loss": 1.3317,
+      "step": 5777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6128960847854614,
+      "learning_rate": 4.8475284166005816e-05,
+      "loss": 1.4353,
+      "step": 5778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9159454107284546,
+      "learning_rate": 4.847501982553529e-05,
+      "loss": 1.5277,
+      "step": 5779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9500240087509155,
+      "learning_rate": 4.847475548506476e-05,
+      "loss": 1.469,
+      "step": 5780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.086463689804077,
+      "learning_rate": 4.8474491144594243e-05,
+      "loss": 1.3764,
+      "step": 5781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4105184078216553,
+      "learning_rate": 4.847422680412371e-05,
+      "loss": 1.3427,
+      "step": 5782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.951155424118042,
+      "learning_rate": 4.847396246365319e-05,
+      "loss": 1.3664,
+      "step": 5783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.659027099609375,
+      "learning_rate": 4.847369812318266e-05,
+      "loss": 2.5036,
+      "step": 5784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0466020107269287,
+      "learning_rate": 4.847343378271214e-05,
+      "loss": 1.2925,
+      "step": 5785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7419512271881104,
+      "learning_rate": 4.847316944224161e-05,
+      "loss": 1.3941,
+      "step": 5786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0035369396209717,
+      "learning_rate": 4.8472905101771085e-05,
+      "loss": 1.3521,
+      "step": 5787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7708044052124023,
+      "learning_rate": 4.847264076130056e-05,
+      "loss": 1.4275,
+      "step": 5788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6456718444824219,
+      "learning_rate": 4.847237642083003e-05,
+      "loss": 1.4782,
+      "step": 5789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7275094985961914,
+      "learning_rate": 4.8472112080359506e-05,
+      "loss": 1.4311,
+      "step": 5790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.084305763244629,
+      "learning_rate": 4.847184773988898e-05,
+      "loss": 1.2888,
+      "step": 5791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8087894916534424,
+      "learning_rate": 4.847158339941845e-05,
+      "loss": 1.4074,
+      "step": 5792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8441691398620605,
+      "learning_rate": 4.8471319058947926e-05,
+      "loss": 1.5027,
+      "step": 5793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4987540245056152,
+      "learning_rate": 4.84710547184774e-05,
+      "loss": 1.4044,
+      "step": 5794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2913873195648193,
+      "learning_rate": 4.8470790378006873e-05,
+      "loss": 1.2456,
+      "step": 5795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.741636037826538,
+      "learning_rate": 4.8470526037536354e-05,
+      "loss": 1.4179,
+      "step": 5796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.670432448387146,
+      "learning_rate": 4.847026169706582e-05,
+      "loss": 1.4609,
+      "step": 5797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6480902433395386,
+      "learning_rate": 4.84699973565953e-05,
+      "loss": 1.3766,
+      "step": 5798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8712096214294434,
+      "learning_rate": 4.846973301612477e-05,
+      "loss": 1.4481,
+      "step": 5799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.728627324104309,
+      "learning_rate": 4.846946867565425e-05,
+      "loss": 1.3729,
+      "step": 5800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2231179475784302,
+      "learning_rate": 4.846920433518372e-05,
+      "loss": 1.3038,
+      "step": 5801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.740309715270996,
+      "learning_rate": 4.846893999471319e-05,
+      "loss": 1.3813,
+      "step": 5802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7007970809936523,
+      "learning_rate": 4.846867565424267e-05,
+      "loss": 1.4004,
+      "step": 5803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4783295392990112,
+      "learning_rate": 4.8468411313772136e-05,
+      "loss": 1.3697,
+      "step": 5804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4422839879989624,
+      "learning_rate": 4.8468146973301616e-05,
+      "loss": 1.2165,
+      "step": 5805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.657866358757019,
+      "learning_rate": 4.846788263283109e-05,
+      "loss": 1.3546,
+      "step": 5806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.405428171157837,
+      "learning_rate": 4.846761829236056e-05,
+      "loss": 2.4269,
+      "step": 5807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.669014811515808,
+      "learning_rate": 4.846735395189004e-05,
+      "loss": 1.3918,
+      "step": 5808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0895321369171143,
+      "learning_rate": 4.846708961141951e-05,
+      "loss": 1.3545,
+      "step": 5809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6673814058303833,
+      "learning_rate": 4.8466825270948984e-05,
+      "loss": 1.3806,
+      "step": 5810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8655140399932861,
+      "learning_rate": 4.846656093047846e-05,
+      "loss": 1.4414,
+      "step": 5811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9546693563461304,
+      "learning_rate": 4.846629659000793e-05,
+      "loss": 1.3703,
+      "step": 5812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8298178911209106,
+      "learning_rate": 4.8466032249537405e-05,
+      "loss": 1.548,
+      "step": 5813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3789175748825073,
+      "learning_rate": 4.846576790906688e-05,
+      "loss": 1.4092,
+      "step": 5814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9434020519256592,
+      "learning_rate": 4.846550356859635e-05,
+      "loss": 1.526,
+      "step": 5815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6064894199371338,
+      "learning_rate": 4.846523922812583e-05,
+      "loss": 1.2967,
+      "step": 5816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5129868984222412,
+      "learning_rate": 4.84649748876553e-05,
+      "loss": 1.2849,
+      "step": 5817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6419628858566284,
+      "learning_rate": 4.846471054718478e-05,
+      "loss": 1.5906,
+      "step": 5818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.273922920227051,
+      "learning_rate": 4.8464446206714246e-05,
+      "loss": 1.4878,
+      "step": 5819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6664180755615234,
+      "learning_rate": 4.8464181866243726e-05,
+      "loss": 1.3987,
+      "step": 5820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6273417472839355,
+      "learning_rate": 4.84639175257732e-05,
+      "loss": 1.3572,
+      "step": 5821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0623226165771484,
+      "learning_rate": 4.8463653185302673e-05,
+      "loss": 1.3504,
+      "step": 5822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0692367553710938,
+      "learning_rate": 4.846338884483215e-05,
+      "loss": 1.1613,
+      "step": 5823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.667441964149475,
+      "learning_rate": 4.846312450436162e-05,
+      "loss": 1.4314,
+      "step": 5824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.097357749938965,
+      "learning_rate": 4.8462860163891094e-05,
+      "loss": 1.4768,
+      "step": 5825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8741570711135864,
+      "learning_rate": 4.846259582342057e-05,
+      "loss": 1.4994,
+      "step": 5826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6883125305175781,
+      "learning_rate": 4.846233148295004e-05,
+      "loss": 1.4016,
+      "step": 5827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9867337942123413,
+      "learning_rate": 4.8462067142479515e-05,
+      "loss": 1.3434,
+      "step": 5828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.273364543914795,
+      "learning_rate": 4.846180280200899e-05,
+      "loss": 1.3931,
+      "step": 5829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9684350490570068,
+      "learning_rate": 4.846153846153846e-05,
+      "loss": 1.4749,
+      "step": 5830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8837523460388184,
+      "learning_rate": 4.846127412106794e-05,
+      "loss": 1.2792,
+      "step": 5831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0115814208984375,
+      "learning_rate": 4.846100978059741e-05,
+      "loss": 1.3326,
+      "step": 5832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7581549882888794,
+      "learning_rate": 4.846074544012689e-05,
+      "loss": 1.498,
+      "step": 5833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7433048486709595,
+      "learning_rate": 4.8460481099656356e-05,
+      "loss": 1.6905,
+      "step": 5834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9508113861083984,
+      "learning_rate": 4.846021675918584e-05,
+      "loss": 1.4126,
+      "step": 5835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4732614755630493,
+      "learning_rate": 4.845995241871531e-05,
+      "loss": 1.4504,
+      "step": 5836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7602614164352417,
+      "learning_rate": 4.8459688078244784e-05,
+      "loss": 1.4088,
+      "step": 5837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8655458688735962,
+      "learning_rate": 4.845942373777426e-05,
+      "loss": 1.4613,
+      "step": 5838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9372882843017578,
+      "learning_rate": 4.8459159397303724e-05,
+      "loss": 1.3006,
+      "step": 5839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6625134944915771,
+      "learning_rate": 4.8458895056833204e-05,
+      "loss": 1.4574,
+      "step": 5840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8654860258102417,
+      "learning_rate": 4.845863071636268e-05,
+      "loss": 1.4256,
+      "step": 5841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8854446411132812,
+      "learning_rate": 4.845836637589215e-05,
+      "loss": 1.456,
+      "step": 5842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0440988540649414,
+      "learning_rate": 4.8458102035421625e-05,
+      "loss": 1.5126,
+      "step": 5843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7498095035552979,
+      "learning_rate": 4.84578376949511e-05,
+      "loss": 1.7421,
+      "step": 5844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7093888521194458,
+      "learning_rate": 4.845757335448057e-05,
+      "loss": 1.4708,
+      "step": 5845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.627274751663208,
+      "learning_rate": 4.845730901401005e-05,
+      "loss": 1.3571,
+      "step": 5846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9190155267715454,
+      "learning_rate": 4.845704467353952e-05,
+      "loss": 1.432,
+      "step": 5847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.632649302482605,
+      "learning_rate": 4.845678033306899e-05,
+      "loss": 1.3966,
+      "step": 5848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4733606576919556,
+      "learning_rate": 4.8456515992598467e-05,
+      "loss": 1.4373,
+      "step": 5849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2430336475372314,
+      "learning_rate": 4.845625165212794e-05,
+      "loss": 1.6044,
+      "step": 5850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9685503244400024,
+      "learning_rate": 4.845598731165742e-05,
+      "loss": 1.404,
+      "step": 5851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.767733335494995,
+      "learning_rate": 4.845572297118689e-05,
+      "loss": 1.517,
+      "step": 5852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0710103511810303,
+      "learning_rate": 4.845545863071637e-05,
+      "loss": 1.5619,
+      "step": 5853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3937467336654663,
+      "learning_rate": 4.8455194290245834e-05,
+      "loss": 1.1472,
+      "step": 5854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9102089405059814,
+      "learning_rate": 4.8454929949775315e-05,
+      "loss": 2.4468,
+      "step": 5855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6123706102371216,
+      "learning_rate": 4.845466560930479e-05,
+      "loss": 1.34,
+      "step": 5856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0462327003479004,
+      "learning_rate": 4.845440126883426e-05,
+      "loss": 1.498,
+      "step": 5857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8184754848480225,
+      "learning_rate": 4.8454136928363736e-05,
+      "loss": 1.3324,
+      "step": 5858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.073859214782715,
+      "learning_rate": 4.845387258789321e-05,
+      "loss": 1.4169,
+      "step": 5859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5116585493087769,
+      "learning_rate": 4.845360824742268e-05,
+      "loss": 1.4489,
+      "step": 5860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5543155670166016,
+      "learning_rate": 4.8453343906952156e-05,
+      "loss": 1.3515,
+      "step": 5861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3851184844970703,
+      "learning_rate": 4.845307956648163e-05,
+      "loss": 1.3392,
+      "step": 5862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7842742204666138,
+      "learning_rate": 4.84528152260111e-05,
+      "loss": 1.3307,
+      "step": 5863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9029113054275513,
+      "learning_rate": 4.845255088554058e-05,
+      "loss": 1.3407,
+      "step": 5864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4590872526168823,
+      "learning_rate": 4.845228654507005e-05,
+      "loss": 1.1881,
+      "step": 5865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2726688385009766,
+      "learning_rate": 4.845202220459953e-05,
+      "loss": 1.6077,
+      "step": 5866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3469276428222656,
+      "learning_rate": 4.8451757864129e-05,
+      "loss": 1.6096,
+      "step": 5867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2330145835876465,
+      "learning_rate": 4.845149352365848e-05,
+      "loss": 1.4564,
+      "step": 5868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9485176801681519,
+      "learning_rate": 4.8451229183187945e-05,
+      "loss": 1.38,
+      "step": 5869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8051180839538574,
+      "learning_rate": 4.8450964842717425e-05,
+      "loss": 1.6673,
+      "step": 5870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.687146782875061,
+      "learning_rate": 4.84507005022469e-05,
+      "loss": 1.5699,
+      "step": 5871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8564085960388184,
+      "learning_rate": 4.845043616177637e-05,
+      "loss": 1.471,
+      "step": 5872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.121131420135498,
+      "learning_rate": 4.8450171821305846e-05,
+      "loss": 1.5568,
+      "step": 5873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6344560384750366,
+      "learning_rate": 4.844990748083531e-05,
+      "loss": 1.3823,
+      "step": 5874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6246540546417236,
+      "learning_rate": 4.844964314036479e-05,
+      "loss": 1.432,
+      "step": 5875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9106330871582031,
+      "learning_rate": 4.8449378799894267e-05,
+      "loss": 1.3027,
+      "step": 5876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.003344774246216,
+      "learning_rate": 4.844911445942374e-05,
+      "loss": 1.6118,
+      "step": 5877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7215330600738525,
+      "learning_rate": 4.8448850118953214e-05,
+      "loss": 1.5174,
+      "step": 5878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.391453981399536,
+      "learning_rate": 4.844858577848269e-05,
+      "loss": 1.6475,
+      "step": 5879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8997163772583008,
+      "learning_rate": 4.844832143801216e-05,
+      "loss": 1.6182,
+      "step": 5880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7039542198181152,
+      "learning_rate": 4.844805709754164e-05,
+      "loss": 1.5044,
+      "step": 5881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8811577558517456,
+      "learning_rate": 4.844779275707111e-05,
+      "loss": 1.4446,
+      "step": 5882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3956289291381836,
+      "learning_rate": 4.844752841660058e-05,
+      "loss": 1.5574,
+      "step": 5883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3624250888824463,
+      "learning_rate": 4.8447264076130055e-05,
+      "loss": 1.2416,
+      "step": 5884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6046565771102905,
+      "learning_rate": 4.844699973565953e-05,
+      "loss": 1.3419,
+      "step": 5885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.944815993309021,
+      "learning_rate": 4.844673539518901e-05,
+      "loss": 1.6677,
+      "step": 5886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.160196304321289,
+      "learning_rate": 4.8446471054718476e-05,
+      "loss": 1.3081,
+      "step": 5887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5366212129592896,
+      "learning_rate": 4.8446206714247956e-05,
+      "loss": 1.2774,
+      "step": 5888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.416046142578125,
+      "learning_rate": 4.844594237377742e-05,
+      "loss": 1.3334,
+      "step": 5889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5525312423706055,
+      "learning_rate": 4.84456780333069e-05,
+      "loss": 1.437,
+      "step": 5890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9905444383621216,
+      "learning_rate": 4.844541369283638e-05,
+      "loss": 1.4552,
+      "step": 5891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5106396675109863,
+      "learning_rate": 4.844514935236585e-05,
+      "loss": 1.2397,
+      "step": 5892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6856703758239746,
+      "learning_rate": 4.8444885011895324e-05,
+      "loss": 1.3984,
+      "step": 5893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1595256328582764,
+      "learning_rate": 4.84446206714248e-05,
+      "loss": 1.5292,
+      "step": 5894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5561280250549316,
+      "learning_rate": 4.844435633095427e-05,
+      "loss": 1.4302,
+      "step": 5895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8560912609100342,
+      "learning_rate": 4.8444091990483745e-05,
+      "loss": 1.2895,
+      "step": 5896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.23507022857666,
+      "learning_rate": 4.844382765001322e-05,
+      "loss": 1.4964,
+      "step": 5897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7972663640975952,
+      "learning_rate": 4.844356330954269e-05,
+      "loss": 1.4504,
+      "step": 5898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.694052815437317,
+      "learning_rate": 4.8443298969072165e-05,
+      "loss": 1.4755,
+      "step": 5899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.904259204864502,
+      "learning_rate": 4.844303462860164e-05,
+      "loss": 1.5028,
+      "step": 5900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6803570985794067,
+      "learning_rate": 4.844277028813112e-05,
+      "loss": 1.309,
+      "step": 5901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2771077156066895,
+      "learning_rate": 4.8442505947660586e-05,
+      "loss": 1.3014,
+      "step": 5902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4967739582061768,
+      "learning_rate": 4.8442241607190067e-05,
+      "loss": 1.4279,
+      "step": 5903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.650825023651123,
+      "learning_rate": 4.844197726671953e-05,
+      "loss": 1.4689,
+      "step": 5904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8678512573242188,
+      "learning_rate": 4.8441712926249014e-05,
+      "loss": 1.4339,
+      "step": 5905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9866645336151123,
+      "learning_rate": 4.844144858577849e-05,
+      "loss": 1.4048,
+      "step": 5906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2837443351745605,
+      "learning_rate": 4.844118424530796e-05,
+      "loss": 1.4651,
+      "step": 5907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4387606382369995,
+      "learning_rate": 4.8440919904837434e-05,
+      "loss": 1.3533,
+      "step": 5908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8451519012451172,
+      "learning_rate": 4.844065556436691e-05,
+      "loss": 1.4015,
+      "step": 5909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8380708694458008,
+      "learning_rate": 4.844039122389638e-05,
+      "loss": 1.5241,
+      "step": 5910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6567280292510986,
+      "learning_rate": 4.8440126883425855e-05,
+      "loss": 1.2168,
+      "step": 5911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7889291048049927,
+      "learning_rate": 4.843986254295533e-05,
+      "loss": 1.543,
+      "step": 5912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6301968097686768,
+      "learning_rate": 4.84395982024848e-05,
+      "loss": 1.4179,
+      "step": 5913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.620779037475586,
+      "learning_rate": 4.8439333862014276e-05,
+      "loss": 1.4693,
+      "step": 5914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8740652799606323,
+      "learning_rate": 4.843906952154375e-05,
+      "loss": 1.3914,
+      "step": 5915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3614580631256104,
+      "learning_rate": 4.843880518107323e-05,
+      "loss": 1.4936,
+      "step": 5916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.656532645225525,
+      "learning_rate": 4.8438540840602697e-05,
+      "loss": 1.3215,
+      "step": 5917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.490763783454895,
+      "learning_rate": 4.843827650013218e-05,
+      "loss": 1.4234,
+      "step": 5918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.163254976272583,
+      "learning_rate": 4.8438012159661644e-05,
+      "loss": 1.4279,
+      "step": 5919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4765214920043945,
+      "learning_rate": 4.843774781919112e-05,
+      "loss": 1.2561,
+      "step": 5920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8741549253463745,
+      "learning_rate": 4.84374834787206e-05,
+      "loss": 1.5546,
+      "step": 5921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.028031349182129,
+      "learning_rate": 4.8437219138250064e-05,
+      "loss": 1.3893,
+      "step": 5922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4210158586502075,
+      "learning_rate": 4.8436954797779545e-05,
+      "loss": 1.358,
+      "step": 5923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6778885126113892,
+      "learning_rate": 4.843669045730901e-05,
+      "loss": 1.5052,
+      "step": 5924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4740965366363525,
+      "learning_rate": 4.843642611683849e-05,
+      "loss": 1.3665,
+      "step": 5925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2669918537139893,
+      "learning_rate": 4.8436161776367965e-05,
+      "loss": 1.3816,
+      "step": 5926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.548175573348999,
+      "learning_rate": 4.843589743589744e-05,
+      "loss": 1.3005,
+      "step": 5927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7350268363952637,
+      "learning_rate": 4.843563309542691e-05,
+      "loss": 1.7632,
+      "step": 5928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9523999691009521,
+      "learning_rate": 4.8435368754956386e-05,
+      "loss": 1.4441,
+      "step": 5929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.531076192855835,
+      "learning_rate": 4.843510441448586e-05,
+      "loss": 1.3875,
+      "step": 5930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6509863138198853,
+      "learning_rate": 4.843484007401533e-05,
+      "loss": 1.3476,
+      "step": 5931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0002591609954834,
+      "learning_rate": 4.843457573354481e-05,
+      "loss": 1.4302,
+      "step": 5932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.98468816280365,
+      "learning_rate": 4.843431139307428e-05,
+      "loss": 1.4726,
+      "step": 5933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7175580263137817,
+      "learning_rate": 4.8434047052603754e-05,
+      "loss": 1.4063,
+      "step": 5934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6701862812042236,
+      "learning_rate": 4.843378271213323e-05,
+      "loss": 1.4062,
+      "step": 5935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.36407208442688,
+      "learning_rate": 4.843351837166271e-05,
+      "loss": 1.5122,
+      "step": 5936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7144834995269775,
+      "learning_rate": 4.8433254031192175e-05,
+      "loss": 1.4533,
+      "step": 5937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8009378910064697,
+      "learning_rate": 4.8432989690721655e-05,
+      "loss": 1.4271,
+      "step": 5938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3714807033538818,
+      "learning_rate": 4.843272535025112e-05,
+      "loss": 1.3273,
+      "step": 5939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5954347848892212,
+      "learning_rate": 4.84324610097806e-05,
+      "loss": 1.4889,
+      "step": 5940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5843819379806519,
+      "learning_rate": 4.8432196669310076e-05,
+      "loss": 1.3273,
+      "step": 5941
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5500154495239258,
+      "learning_rate": 4.843193232883955e-05,
+      "loss": 1.2931,
+      "step": 5942
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8264559507369995,
+      "learning_rate": 4.843166798836902e-05,
+      "loss": 1.4039,
+      "step": 5943
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0670228004455566,
+      "learning_rate": 4.8431403647898497e-05,
+      "loss": 1.4263,
+      "step": 5944
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.341623306274414,
+      "learning_rate": 4.843113930742797e-05,
+      "loss": 1.2882,
+      "step": 5945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7338883876800537,
+      "learning_rate": 4.8430874966957444e-05,
+      "loss": 1.3354,
+      "step": 5946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6274378299713135,
+      "learning_rate": 4.843061062648692e-05,
+      "loss": 1.3741,
+      "step": 5947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1950230598449707,
+      "learning_rate": 4.843034628601639e-05,
+      "loss": 1.4118,
+      "step": 5948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4571107625961304,
+      "learning_rate": 4.8430081945545864e-05,
+      "loss": 1.4463,
+      "step": 5949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7952215671539307,
+      "learning_rate": 4.842981760507534e-05,
+      "loss": 1.4562,
+      "step": 5950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7082561254501343,
+      "learning_rate": 4.842955326460482e-05,
+      "loss": 1.4023,
+      "step": 5951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.339012861251831,
+      "learning_rate": 4.8429288924134285e-05,
+      "loss": 1.4332,
+      "step": 5952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.029994249343872,
+      "learning_rate": 4.8429024583663765e-05,
+      "loss": 1.4643,
+      "step": 5953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4674406051635742,
+      "learning_rate": 4.842876024319323e-05,
+      "loss": 1.136,
+      "step": 5954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.136595726013184,
+      "learning_rate": 4.842849590272271e-05,
+      "loss": 1.7012,
+      "step": 5955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0533084869384766,
+      "learning_rate": 4.8428231562252186e-05,
+      "loss": 1.4322,
+      "step": 5956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.721919298171997,
+      "learning_rate": 4.842796722178165e-05,
+      "loss": 1.4209,
+      "step": 5957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5380709171295166,
+      "learning_rate": 4.842770288131113e-05,
+      "loss": 1.4523,
+      "step": 5958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.20076060295105,
+      "learning_rate": 4.84274385408406e-05,
+      "loss": 1.6729,
+      "step": 5959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9844956398010254,
+      "learning_rate": 4.842717420037008e-05,
+      "loss": 1.3661,
+      "step": 5960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7559512853622437,
+      "learning_rate": 4.8426909859899554e-05,
+      "loss": 1.3216,
+      "step": 5961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.738481044769287,
+      "learning_rate": 4.842664551942903e-05,
+      "loss": 1.3749,
+      "step": 5962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7872519493103027,
+      "learning_rate": 4.84263811789585e-05,
+      "loss": 1.416,
+      "step": 5963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8900316953659058,
+      "learning_rate": 4.8426116838487975e-05,
+      "loss": 1.3967,
+      "step": 5964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1465349197387695,
+      "learning_rate": 4.842585249801745e-05,
+      "loss": 1.5195,
+      "step": 5965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8195936679840088,
+      "learning_rate": 4.842558815754692e-05,
+      "loss": 1.4036,
+      "step": 5966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.024341583251953,
+      "learning_rate": 4.8425323817076395e-05,
+      "loss": 1.3097,
+      "step": 5967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.836979866027832,
+      "learning_rate": 4.842505947660587e-05,
+      "loss": 1.3867,
+      "step": 5968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8989709615707397,
+      "learning_rate": 4.842479513613534e-05,
+      "loss": 1.5386,
+      "step": 5969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0845768451690674,
+      "learning_rate": 4.8424530795664816e-05,
+      "loss": 1.5598,
+      "step": 5970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3178584575653076,
+      "learning_rate": 4.8424266455194296e-05,
+      "loss": 1.5134,
+      "step": 5971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.188380002975464,
+      "learning_rate": 4.842400211472376e-05,
+      "loss": 1.4588,
+      "step": 5972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0621747970581055,
+      "learning_rate": 4.8423737774253244e-05,
+      "loss": 1.4154,
+      "step": 5973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5316870212554932,
+      "learning_rate": 4.842347343378271e-05,
+      "loss": 1.2077,
+      "step": 5974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.82865571975708,
+      "learning_rate": 4.842320909331219e-05,
+      "loss": 1.6106,
+      "step": 5975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.021101236343384,
+      "learning_rate": 4.8422944752841664e-05,
+      "loss": 1.456,
+      "step": 5976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4527124166488647,
+      "learning_rate": 4.842268041237114e-05,
+      "loss": 1.2343,
+      "step": 5977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8262113332748413,
+      "learning_rate": 4.842241607190061e-05,
+      "loss": 1.296,
+      "step": 5978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.500457525253296,
+      "learning_rate": 4.8422151731430085e-05,
+      "loss": 2.1147,
+      "step": 5979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8740986585617065,
+      "learning_rate": 4.842188739095956e-05,
+      "loss": 1.3834,
+      "step": 5980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7808442115783691,
+      "learning_rate": 4.842162305048903e-05,
+      "loss": 1.378,
+      "step": 5981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7709171772003174,
+      "learning_rate": 4.8421358710018506e-05,
+      "loss": 1.415,
+      "step": 5982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8967705965042114,
+      "learning_rate": 4.842109436954798e-05,
+      "loss": 1.4138,
+      "step": 5983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4771448373794556,
+      "learning_rate": 4.842083002907745e-05,
+      "loss": 1.4341,
+      "step": 5984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.273117780685425,
+      "learning_rate": 4.8420565688606926e-05,
+      "loss": 1.547,
+      "step": 5985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2445123195648193,
+      "learning_rate": 4.842030134813641e-05,
+      "loss": 1.521,
+      "step": 5986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9863500595092773,
+      "learning_rate": 4.8420037007665874e-05,
+      "loss": 1.4018,
+      "step": 5987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.080561637878418,
+      "learning_rate": 4.8419772667195354e-05,
+      "loss": 1.428,
+      "step": 5988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.02518892288208,
+      "learning_rate": 4.841950832672482e-05,
+      "loss": 1.4618,
+      "step": 5989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.036107301712036,
+      "learning_rate": 4.84192439862543e-05,
+      "loss": 1.6507,
+      "step": 5990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2027151584625244,
+      "learning_rate": 4.8418979645783775e-05,
+      "loss": 1.3569,
+      "step": 5991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5087175369262695,
+      "learning_rate": 4.841871530531324e-05,
+      "loss": 1.2286,
+      "step": 5992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.774732232093811,
+      "learning_rate": 4.841845096484272e-05,
+      "loss": 1.4937,
+      "step": 5993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9310197830200195,
+      "learning_rate": 4.841818662437219e-05,
+      "loss": 1.4071,
+      "step": 5994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3090226650238037,
+      "learning_rate": 4.841792228390167e-05,
+      "loss": 1.9078,
+      "step": 5995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5815956592559814,
+      "learning_rate": 4.841765794343114e-05,
+      "loss": 1.358,
+      "step": 5996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4588104486465454,
+      "learning_rate": 4.8417393602960616e-05,
+      "loss": 1.3207,
+      "step": 5997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.801145076751709,
+      "learning_rate": 4.841712926249009e-05,
+      "loss": 1.413,
+      "step": 5998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.803552269935608,
+      "learning_rate": 4.841686492201956e-05,
+      "loss": 1.3751,
+      "step": 5999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5930027961730957,
+      "learning_rate": 4.841660058154904e-05,
+      "loss": 1.368,
+      "step": 6000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4202892780303955,
+      "learning_rate": 4.841633624107851e-05,
+      "loss": 1.3787,
+      "step": 6001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9479396343231201,
+      "learning_rate": 4.8416071900607984e-05,
+      "loss": 1.4622,
+      "step": 6002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.568363904953003,
+      "learning_rate": 4.841580756013746e-05,
+      "loss": 1.4183,
+      "step": 6003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9793590307235718,
+      "learning_rate": 4.841554321966693e-05,
+      "loss": 1.5965,
+      "step": 6004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.217047929763794,
+      "learning_rate": 4.8415278879196405e-05,
+      "loss": 1.6381,
+      "step": 6005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6210744380950928,
+      "learning_rate": 4.8415014538725885e-05,
+      "loss": 1.3212,
+      "step": 6006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7058119773864746,
+      "learning_rate": 4.841475019825535e-05,
+      "loss": 1.4066,
+      "step": 6007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6952123641967773,
+      "learning_rate": 4.841448585778483e-05,
+      "loss": 1.477,
+      "step": 6008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5785906314849854,
+      "learning_rate": 4.84142215173143e-05,
+      "loss": 2.0974,
+      "step": 6009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8473061323165894,
+      "learning_rate": 4.841395717684378e-05,
+      "loss": 1.3532,
+      "step": 6010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5085325241088867,
+      "learning_rate": 4.841369283637325e-05,
+      "loss": 1.5058,
+      "step": 6011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7637439966201782,
+      "learning_rate": 4.8413428495902726e-05,
+      "loss": 1.3937,
+      "step": 6012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5932888984680176,
+      "learning_rate": 4.84131641554322e-05,
+      "loss": 1.3226,
+      "step": 6013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.093410015106201,
+      "learning_rate": 4.8412899814961674e-05,
+      "loss": 1.5131,
+      "step": 6014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7985966205596924,
+      "learning_rate": 4.841263547449115e-05,
+      "loss": 1.3396,
+      "step": 6015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.625970721244812,
+      "learning_rate": 4.841237113402062e-05,
+      "loss": 1.7793,
+      "step": 6016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0310299396514893,
+      "learning_rate": 4.8412106793550094e-05,
+      "loss": 1.5702,
+      "step": 6017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7560282945632935,
+      "learning_rate": 4.841184245307957e-05,
+      "loss": 1.2684,
+      "step": 6018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9292409420013428,
+      "learning_rate": 4.841157811260904e-05,
+      "loss": 1.569,
+      "step": 6019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3320858478546143,
+      "learning_rate": 4.8411313772138515e-05,
+      "loss": 1.2546,
+      "step": 6020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.830118179321289,
+      "learning_rate": 4.8411049431667995e-05,
+      "loss": 1.9694,
+      "step": 6021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9703755378723145,
+      "learning_rate": 4.841078509119746e-05,
+      "loss": 1.4686,
+      "step": 6022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1245367527008057,
+      "learning_rate": 4.841052075072694e-05,
+      "loss": 1.4907,
+      "step": 6023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7603442668914795,
+      "learning_rate": 4.841025641025641e-05,
+      "loss": 1.6045,
+      "step": 6024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5304856300354004,
+      "learning_rate": 4.840999206978589e-05,
+      "loss": 1.8548,
+      "step": 6025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.584293246269226,
+      "learning_rate": 4.840972772931536e-05,
+      "loss": 2.4767,
+      "step": 6026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7177733182907104,
+      "learning_rate": 4.840946338884484e-05,
+      "loss": 1.4386,
+      "step": 6027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7882304191589355,
+      "learning_rate": 4.840919904837431e-05,
+      "loss": 1.4933,
+      "step": 6028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5551881790161133,
+      "learning_rate": 4.840893470790378e-05,
+      "loss": 1.4046,
+      "step": 6029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2046337127685547,
+      "learning_rate": 4.840867036743326e-05,
+      "loss": 1.5933,
+      "step": 6030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7296518087387085,
+      "learning_rate": 4.840840602696273e-05,
+      "loss": 1.4311,
+      "step": 6031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0331151485443115,
+      "learning_rate": 4.8408141686492205e-05,
+      "loss": 1.4183,
+      "step": 6032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4345885515213013,
+      "learning_rate": 4.840787734602168e-05,
+      "loss": 1.2904,
+      "step": 6033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9695258140563965,
+      "learning_rate": 4.840761300555115e-05,
+      "loss": 1.4764,
+      "step": 6034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9519201517105103,
+      "learning_rate": 4.8407348665080625e-05,
+      "loss": 1.4024,
+      "step": 6035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9538984298706055,
+      "learning_rate": 4.8407084324610106e-05,
+      "loss": 1.4755,
+      "step": 6036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7535780668258667,
+      "learning_rate": 4.840681998413957e-05,
+      "loss": 1.4871,
+      "step": 6037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9031850099563599,
+      "learning_rate": 4.8406555643669046e-05,
+      "loss": 1.7884,
+      "step": 6038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1076838970184326,
+      "learning_rate": 4.840629130319852e-05,
+      "loss": 1.5342,
+      "step": 6039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.216662883758545,
+      "learning_rate": 4.840602696272799e-05,
+      "loss": 1.6811,
+      "step": 6040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8524043560028076,
+      "learning_rate": 4.8405762622257474e-05,
+      "loss": 1.4456,
+      "step": 6041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5259125232696533,
+      "learning_rate": 4.840549828178694e-05,
+      "loss": 2.3605,
+      "step": 6042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7552589178085327,
+      "learning_rate": 4.840523394131642e-05,
+      "loss": 1.3622,
+      "step": 6043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.981697678565979,
+      "learning_rate": 4.840496960084589e-05,
+      "loss": 1.4931,
+      "step": 6044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8973721265792847,
+      "learning_rate": 4.840470526037537e-05,
+      "loss": 1.5377,
+      "step": 6045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8032373189926147,
+      "learning_rate": 4.840444091990484e-05,
+      "loss": 1.5846,
+      "step": 6046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.672732949256897,
+      "learning_rate": 4.8404176579434315e-05,
+      "loss": 1.4026,
+      "step": 6047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9174094200134277,
+      "learning_rate": 4.840391223896379e-05,
+      "loss": 1.4225,
+      "step": 6048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6702638864517212,
+      "learning_rate": 4.840364789849326e-05,
+      "loss": 1.4549,
+      "step": 6049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6555778980255127,
+      "learning_rate": 4.8403383558022736e-05,
+      "loss": 1.4135,
+      "step": 6050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7418198585510254,
+      "learning_rate": 4.840311921755221e-05,
+      "loss": 1.492,
+      "step": 6051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5808851718902588,
+      "learning_rate": 4.840285487708168e-05,
+      "loss": 1.3557,
+      "step": 6052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9051976203918457,
+      "learning_rate": 4.8402590536611156e-05,
+      "loss": 1.5893,
+      "step": 6053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8803669214248657,
+      "learning_rate": 4.840232619614063e-05,
+      "loss": 1.4468,
+      "step": 6054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7576953172683716,
+      "learning_rate": 4.8402061855670104e-05,
+      "loss": 1.4247,
+      "step": 6055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1546995639801025,
+      "learning_rate": 4.8401797515199584e-05,
+      "loss": 1.391,
+      "step": 6056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2778875827789307,
+      "learning_rate": 4.840153317472905e-05,
+      "loss": 1.5754,
+      "step": 6057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9502428770065308,
+      "learning_rate": 4.840126883425853e-05,
+      "loss": 1.4034,
+      "step": 6058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.875204086303711,
+      "learning_rate": 4.8401004493788e-05,
+      "loss": 1.2696,
+      "step": 6059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.595214605331421,
+      "learning_rate": 4.840074015331748e-05,
+      "loss": 1.4891,
+      "step": 6060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7254564762115479,
+      "learning_rate": 4.840047581284695e-05,
+      "loss": 1.3965,
+      "step": 6061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.504562258720398,
+      "learning_rate": 4.8400211472376425e-05,
+      "loss": 1.4016,
+      "step": 6062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6035069227218628,
+      "learning_rate": 4.83999471319059e-05,
+      "loss": 1.3976,
+      "step": 6063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.632091999053955,
+      "learning_rate": 4.8399682791435366e-05,
+      "loss": 1.3236,
+      "step": 6064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.414618730545044,
+      "learning_rate": 4.8399418450964846e-05,
+      "loss": 1.4519,
+      "step": 6065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6097583770751953,
+      "learning_rate": 4.839915411049432e-05,
+      "loss": 1.4599,
+      "step": 6066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1019625663757324,
+      "learning_rate": 4.839888977002379e-05,
+      "loss": 1.9564,
+      "step": 6067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.804720878601074,
+      "learning_rate": 4.839862542955327e-05,
+      "loss": 1.5361,
+      "step": 6068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4190549850463867,
+      "learning_rate": 4.839836108908274e-05,
+      "loss": 1.3691,
+      "step": 6069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5714927911758423,
+      "learning_rate": 4.8398096748612214e-05,
+      "loss": 1.3367,
+      "step": 6070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9285069704055786,
+      "learning_rate": 4.8397832408141694e-05,
+      "loss": 1.4144,
+      "step": 6071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2992210388183594,
+      "learning_rate": 4.839756806767116e-05,
+      "loss": 1.2827,
+      "step": 6072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4958614110946655,
+      "learning_rate": 4.839730372720064e-05,
+      "loss": 1.258,
+      "step": 6073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.628309965133667,
+      "learning_rate": 4.839703938673011e-05,
+      "loss": 1.4581,
+      "step": 6074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.013759136199951,
+      "learning_rate": 4.839677504625958e-05,
+      "loss": 1.4638,
+      "step": 6075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1463534832000732,
+      "learning_rate": 4.839651070578906e-05,
+      "loss": 1.601,
+      "step": 6076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7048437595367432,
+      "learning_rate": 4.839624636531853e-05,
+      "loss": 1.4605,
+      "step": 6077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9689546823501587,
+      "learning_rate": 4.839598202484801e-05,
+      "loss": 1.3185,
+      "step": 6078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4981482028961182,
+      "learning_rate": 4.8395717684377476e-05,
+      "loss": 1.272,
+      "step": 6079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.636247992515564,
+      "learning_rate": 4.8395453343906956e-05,
+      "loss": 1.3734,
+      "step": 6080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5858428478240967,
+      "learning_rate": 4.839518900343643e-05,
+      "loss": 1.3651,
+      "step": 6081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8602960109710693,
+      "learning_rate": 4.8394924662965904e-05,
+      "loss": 1.3925,
+      "step": 6082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6307518482208252,
+      "learning_rate": 4.839466032249538e-05,
+      "loss": 1.4083,
+      "step": 6083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8157460689544678,
+      "learning_rate": 4.839439598202485e-05,
+      "loss": 1.4242,
+      "step": 6084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1332736015319824,
+      "learning_rate": 4.8394131641554324e-05,
+      "loss": 1.4055,
+      "step": 6085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4607826471328735,
+      "learning_rate": 4.83938673010838e-05,
+      "loss": 1.2894,
+      "step": 6086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5489532947540283,
+      "learning_rate": 4.839360296061327e-05,
+      "loss": 2.155,
+      "step": 6087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0093133449554443,
+      "learning_rate": 4.8393338620142745e-05,
+      "loss": 1.5408,
+      "step": 6088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7938201427459717,
+      "learning_rate": 4.839307427967222e-05,
+      "loss": 1.4893,
+      "step": 6089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9420911073684692,
+      "learning_rate": 4.839280993920169e-05,
+      "loss": 1.4069,
+      "step": 6090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2374913692474365,
+      "learning_rate": 4.839254559873117e-05,
+      "loss": 1.5566,
+      "step": 6091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7230379581451416,
+      "learning_rate": 4.839228125826064e-05,
+      "loss": 1.3196,
+      "step": 6092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4814276695251465,
+      "learning_rate": 4.839201691779012e-05,
+      "loss": 1.3831,
+      "step": 6093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8867626190185547,
+      "learning_rate": 4.8391752577319586e-05,
+      "loss": 1.5002,
+      "step": 6094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8894315958023071,
+      "learning_rate": 4.839148823684907e-05,
+      "loss": 2.5155,
+      "step": 6095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.288316011428833,
+      "learning_rate": 4.839122389637854e-05,
+      "loss": 1.3271,
+      "step": 6096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5970419645309448,
+      "learning_rate": 4.8390959555908014e-05,
+      "loss": 1.4565,
+      "step": 6097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7331547737121582,
+      "learning_rate": 4.839069521543749e-05,
+      "loss": 1.2268,
+      "step": 6098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4795880317687988,
+      "learning_rate": 4.839043087496696e-05,
+      "loss": 2.2292,
+      "step": 6099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0718016624450684,
+      "learning_rate": 4.8390166534496435e-05,
+      "loss": 1.3608,
+      "step": 6100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7821236848831177,
+      "learning_rate": 4.838990219402591e-05,
+      "loss": 1.4415,
+      "step": 6101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1660101413726807,
+      "learning_rate": 4.838963785355538e-05,
+      "loss": 1.4597,
+      "step": 6102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6507564783096313,
+      "learning_rate": 4.8389373513084855e-05,
+      "loss": 1.4836,
+      "step": 6103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.687912106513977,
+      "learning_rate": 4.838910917261433e-05,
+      "loss": 1.4968,
+      "step": 6104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.284611463546753,
+      "learning_rate": 4.83888448321438e-05,
+      "loss": 1.468,
+      "step": 6105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5659557580947876,
+      "learning_rate": 4.838858049167328e-05,
+      "loss": 1.2681,
+      "step": 6106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8584054708480835,
+      "learning_rate": 4.838831615120275e-05,
+      "loss": 1.3629,
+      "step": 6107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7195202112197876,
+      "learning_rate": 4.838805181073223e-05,
+      "loss": 1.3651,
+      "step": 6108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.428870677947998,
+      "learning_rate": 4.83877874702617e-05,
+      "loss": 1.468,
+      "step": 6109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6679024696350098,
+      "learning_rate": 4.838752312979117e-05,
+      "loss": 1.3551,
+      "step": 6110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3351950645446777,
+      "learning_rate": 4.838725878932065e-05,
+      "loss": 1.214,
+      "step": 6111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0594308376312256,
+      "learning_rate": 4.838699444885012e-05,
+      "loss": 1.4014,
+      "step": 6112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.520538330078125,
+      "learning_rate": 4.83867301083796e-05,
+      "loss": 1.2827,
+      "step": 6113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1707305908203125,
+      "learning_rate": 4.8386465767909065e-05,
+      "loss": 1.4123,
+      "step": 6114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5471241474151611,
+      "learning_rate": 4.8386201427438545e-05,
+      "loss": 1.4366,
+      "step": 6115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7113407850265503,
+      "learning_rate": 4.838593708696802e-05,
+      "loss": 1.5174,
+      "step": 6116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0511794090270996,
+      "learning_rate": 4.838567274649749e-05,
+      "loss": 1.5512,
+      "step": 6117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3596208095550537,
+      "learning_rate": 4.8385408406026966e-05,
+      "loss": 1.5505,
+      "step": 6118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.537919044494629,
+      "learning_rate": 4.838514406555644e-05,
+      "loss": 1.5358,
+      "step": 6119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7392295598983765,
+      "learning_rate": 4.838487972508591e-05,
+      "loss": 1.4316,
+      "step": 6120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6761339902877808,
+      "learning_rate": 4.8384615384615386e-05,
+      "loss": 1.387,
+      "step": 6121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.691961646080017,
+      "learning_rate": 4.838435104414486e-05,
+      "loss": 1.3776,
+      "step": 6122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.176518678665161,
+      "learning_rate": 4.8384086703674333e-05,
+      "loss": 1.502,
+      "step": 6123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7518404722213745,
+      "learning_rate": 4.838382236320381e-05,
+      "loss": 1.5573,
+      "step": 6124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6806093454360962,
+      "learning_rate": 4.838355802273328e-05,
+      "loss": 1.6605,
+      "step": 6125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5381968021392822,
+      "learning_rate": 4.838329368226276e-05,
+      "loss": 1.4262,
+      "step": 6126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8382492065429688,
+      "learning_rate": 4.838302934179223e-05,
+      "loss": 1.4366,
+      "step": 6127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.708070993423462,
+      "learning_rate": 4.838276500132171e-05,
+      "loss": 1.3316,
+      "step": 6128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0687365531921387,
+      "learning_rate": 4.8382500660851175e-05,
+      "loss": 1.6066,
+      "step": 6129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1773464679718018,
+      "learning_rate": 4.8382236320380655e-05,
+      "loss": 1.5109,
+      "step": 6130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7985814809799194,
+      "learning_rate": 4.838197197991013e-05,
+      "loss": 1.4936,
+      "step": 6131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.390382766723633,
+      "learning_rate": 4.83817076394396e-05,
+      "loss": 1.378,
+      "step": 6132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5768065452575684,
+      "learning_rate": 4.8381443298969076e-05,
+      "loss": 1.3719,
+      "step": 6133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8065859079360962,
+      "learning_rate": 4.838117895849855e-05,
+      "loss": 1.3714,
+      "step": 6134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.566085696220398,
+      "learning_rate": 4.838091461802802e-05,
+      "loss": 1.3866,
+      "step": 6135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9985506534576416,
+      "learning_rate": 4.83806502775575e-05,
+      "loss": 1.5219,
+      "step": 6136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8659248352050781,
+      "learning_rate": 4.838038593708697e-05,
+      "loss": 1.445,
+      "step": 6137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9594008922576904,
+      "learning_rate": 4.8380121596616444e-05,
+      "loss": 1.375,
+      "step": 6138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6263427734375,
+      "learning_rate": 4.837985725614592e-05,
+      "loss": 1.3448,
+      "step": 6139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.171999454498291,
+      "learning_rate": 4.837959291567539e-05,
+      "loss": 2.18,
+      "step": 6140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2973012924194336,
+      "learning_rate": 4.837932857520487e-05,
+      "loss": 1.5072,
+      "step": 6141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6555209159851074,
+      "learning_rate": 4.837906423473434e-05,
+      "loss": 1.4839,
+      "step": 6142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9893873929977417,
+      "learning_rate": 4.837879989426382e-05,
+      "loss": 1.4356,
+      "step": 6143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2842626571655273,
+      "learning_rate": 4.8378535553793285e-05,
+      "loss": 1.4983,
+      "step": 6144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.642059326171875,
+      "learning_rate": 4.8378271213322766e-05,
+      "loss": 1.4724,
+      "step": 6145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0432779788970947,
+      "learning_rate": 4.837800687285224e-05,
+      "loss": 1.2883,
+      "step": 6146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.258097171783447,
+      "learning_rate": 4.8377742532381706e-05,
+      "loss": 1.6741,
+      "step": 6147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.916152834892273,
+      "learning_rate": 4.8377478191911186e-05,
+      "loss": 1.3688,
+      "step": 6148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4837087392807007,
+      "learning_rate": 4.837721385144065e-05,
+      "loss": 1.4239,
+      "step": 6149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0862348079681396,
+      "learning_rate": 4.8376949510970133e-05,
+      "loss": 1.4795,
+      "step": 6150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1864593029022217,
+      "learning_rate": 4.837668517049961e-05,
+      "loss": 1.7361,
+      "step": 6151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0611395835876465,
+      "learning_rate": 4.837642083002908e-05,
+      "loss": 1.3795,
+      "step": 6152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9982240200042725,
+      "learning_rate": 4.8376156489558554e-05,
+      "loss": 1.3216,
+      "step": 6153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.719698190689087,
+      "learning_rate": 4.837589214908803e-05,
+      "loss": 1.5873,
+      "step": 6154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.441307306289673,
+      "learning_rate": 4.83756278086175e-05,
+      "loss": 1.5384,
+      "step": 6155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6217659711837769,
+      "learning_rate": 4.8375363468146975e-05,
+      "loss": 1.3541,
+      "step": 6156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2831171751022339,
+      "learning_rate": 4.837509912767645e-05,
+      "loss": 1.1653,
+      "step": 6157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6001358032226562,
+      "learning_rate": 4.837483478720592e-05,
+      "loss": 1.3868,
+      "step": 6158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.569635033607483,
+      "learning_rate": 4.8374570446735396e-05,
+      "loss": 1.3973,
+      "step": 6159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3727385997772217,
+      "learning_rate": 4.837430610626487e-05,
+      "loss": 1.4473,
+      "step": 6160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6972324848175049,
+      "learning_rate": 4.837404176579435e-05,
+      "loss": 1.2894,
+      "step": 6161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1059634685516357,
+      "learning_rate": 4.8373777425323816e-05,
+      "loss": 1.5279,
+      "step": 6162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.968305230140686,
+      "learning_rate": 4.83735130848533e-05,
+      "loss": 1.3402,
+      "step": 6163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5613653659820557,
+      "learning_rate": 4.8373248744382763e-05,
+      "loss": 1.1392,
+      "step": 6164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6044131517410278,
+      "learning_rate": 4.8372984403912244e-05,
+      "loss": 1.4598,
+      "step": 6165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4059150218963623,
+      "learning_rate": 4.837272006344172e-05,
+      "loss": 1.7135,
+      "step": 6166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5571317672729492,
+      "learning_rate": 4.837245572297119e-05,
+      "loss": 1.3129,
+      "step": 6167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.947374701499939,
+      "learning_rate": 4.8372191382500664e-05,
+      "loss": 1.2815,
+      "step": 6168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9128186702728271,
+      "learning_rate": 4.837192704203014e-05,
+      "loss": 1.4633,
+      "step": 6169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7642194032669067,
+      "learning_rate": 4.837166270155961e-05,
+      "loss": 1.403,
+      "step": 6170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7087682485580444,
+      "learning_rate": 4.8371398361089085e-05,
+      "loss": 1.4766,
+      "step": 6171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5011974573135376,
+      "learning_rate": 4.837113402061856e-05,
+      "loss": 1.6115,
+      "step": 6172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6970925331115723,
+      "learning_rate": 4.837086968014803e-05,
+      "loss": 1.292,
+      "step": 6173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5521448850631714,
+      "learning_rate": 4.8370605339677506e-05,
+      "loss": 1.3877,
+      "step": 6174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5931133031845093,
+      "learning_rate": 4.837034099920698e-05,
+      "loss": 1.4663,
+      "step": 6175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.306830883026123,
+      "learning_rate": 4.837007665873646e-05,
+      "loss": 1.6688,
+      "step": 6176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.103497266769409,
+      "learning_rate": 4.836981231826593e-05,
+      "loss": 1.5324,
+      "step": 6177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6159942150115967,
+      "learning_rate": 4.836954797779541e-05,
+      "loss": 1.392,
+      "step": 6178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0038318634033203,
+      "learning_rate": 4.8369283637324874e-05,
+      "loss": 1.4534,
+      "step": 6179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.722639799118042,
+      "learning_rate": 4.8369019296854354e-05,
+      "loss": 1.3258,
+      "step": 6180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6428675651550293,
+      "learning_rate": 4.836875495638383e-05,
+      "loss": 1.3843,
+      "step": 6181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9443061351776123,
+      "learning_rate": 4.8368490615913294e-05,
+      "loss": 1.3521,
+      "step": 6182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3829474449157715,
+      "learning_rate": 4.8368226275442775e-05,
+      "loss": 1.4541,
+      "step": 6183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9254757165908813,
+      "learning_rate": 4.836796193497224e-05,
+      "loss": 1.4291,
+      "step": 6184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.980299234390259,
+      "learning_rate": 4.836769759450172e-05,
+      "loss": 1.5833,
+      "step": 6185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7734400033950806,
+      "learning_rate": 4.8367433254031196e-05,
+      "loss": 1.394,
+      "step": 6186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.332939386367798,
+      "learning_rate": 4.836716891356067e-05,
+      "loss": 1.5466,
+      "step": 6187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7362735271453857,
+      "learning_rate": 4.836690457309014e-05,
+      "loss": 1.6575,
+      "step": 6188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.720354437828064,
+      "learning_rate": 4.8366640232619616e-05,
+      "loss": 1.4604,
+      "step": 6189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6780904531478882,
+      "learning_rate": 4.836637589214909e-05,
+      "loss": 1.5092,
+      "step": 6190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9058243036270142,
+      "learning_rate": 4.836611155167857e-05,
+      "loss": 1.5115,
+      "step": 6191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.476996898651123,
+      "learning_rate": 4.836584721120804e-05,
+      "loss": 1.1531,
+      "step": 6192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8778434991836548,
+      "learning_rate": 4.836558287073751e-05,
+      "loss": 1.5036,
+      "step": 6193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3519941568374634,
+      "learning_rate": 4.8365318530266984e-05,
+      "loss": 1.31,
+      "step": 6194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6786155700683594,
+      "learning_rate": 4.836505418979646e-05,
+      "loss": 1.4009,
+      "step": 6195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.121520757675171,
+      "learning_rate": 4.836478984932594e-05,
+      "loss": 1.4541,
+      "step": 6196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5522239208221436,
+      "learning_rate": 4.8364525508855405e-05,
+      "loss": 2.2565,
+      "step": 6197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2698941230773926,
+      "learning_rate": 4.8364261168384885e-05,
+      "loss": 1.3536,
+      "step": 6198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3673436641693115,
+      "learning_rate": 4.836399682791435e-05,
+      "loss": 1.4209,
+      "step": 6199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6515331268310547,
+      "learning_rate": 4.836373248744383e-05,
+      "loss": 1.4727,
+      "step": 6200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7506335973739624,
+      "learning_rate": 4.8363468146973306e-05,
+      "loss": 1.4915,
+      "step": 6201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0116634368896484,
+      "learning_rate": 4.836320380650278e-05,
+      "loss": 1.3859,
+      "step": 6202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.469986915588379,
+      "learning_rate": 4.836293946603225e-05,
+      "loss": 1.471,
+      "step": 6203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6455501317977905,
+      "learning_rate": 4.8362675125561727e-05,
+      "loss": 1.3971,
+      "step": 6204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.559576153755188,
+      "learning_rate": 4.83624107850912e-05,
+      "loss": 1.3502,
+      "step": 6205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.016934633255005,
+      "learning_rate": 4.8362146444620674e-05,
+      "loss": 1.5411,
+      "step": 6206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.358440399169922,
+      "learning_rate": 4.836188210415015e-05,
+      "loss": 1.5534,
+      "step": 6207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8668150901794434,
+      "learning_rate": 4.836161776367962e-05,
+      "loss": 1.3462,
+      "step": 6208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.799009919166565,
+      "learning_rate": 4.8361353423209094e-05,
+      "loss": 1.4505,
+      "step": 6209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7629196643829346,
+      "learning_rate": 4.836108908273857e-05,
+      "loss": 1.6589,
+      "step": 6210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7770651578903198,
+      "learning_rate": 4.836082474226805e-05,
+      "loss": 1.28,
+      "step": 6211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9307637214660645,
+      "learning_rate": 4.8360560401797515e-05,
+      "loss": 1.364,
+      "step": 6212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5292402505874634,
+      "learning_rate": 4.8360296061326996e-05,
+      "loss": 1.2966,
+      "step": 6213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6868131160736084,
+      "learning_rate": 4.836003172085646e-05,
+      "loss": 1.4575,
+      "step": 6214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.545038104057312,
+      "learning_rate": 4.835976738038594e-05,
+      "loss": 1.3189,
+      "step": 6215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0945894718170166,
+      "learning_rate": 4.8359503039915416e-05,
+      "loss": 1.3478,
+      "step": 6216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2285497188568115,
+      "learning_rate": 4.835923869944489e-05,
+      "loss": 1.474,
+      "step": 6217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5839710235595703,
+      "learning_rate": 4.835897435897436e-05,
+      "loss": 1.6558,
+      "step": 6218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.780615210533142,
+      "learning_rate": 4.835871001850383e-05,
+      "loss": 1.4707,
+      "step": 6219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6688090562820435,
+      "learning_rate": 4.835844567803331e-05,
+      "loss": 1.3316,
+      "step": 6220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0766494274139404,
+      "learning_rate": 4.8358181337562784e-05,
+      "loss": 1.5217,
+      "step": 6221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7954343557357788,
+      "learning_rate": 4.835791699709226e-05,
+      "loss": 1.4549,
+      "step": 6222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8200305700302124,
+      "learning_rate": 4.835765265662173e-05,
+      "loss": 1.4033,
+      "step": 6223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.933725595474243,
+      "learning_rate": 4.8357388316151205e-05,
+      "loss": 1.4823,
+      "step": 6224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.023911476135254,
+      "learning_rate": 4.835712397568068e-05,
+      "loss": 1.4505,
+      "step": 6225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.034367561340332,
+      "learning_rate": 4.835685963521016e-05,
+      "loss": 1.4652,
+      "step": 6226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.554558515548706,
+      "learning_rate": 4.8356595294739625e-05,
+      "loss": 1.4602,
+      "step": 6227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0489308834075928,
+      "learning_rate": 4.83563309542691e-05,
+      "loss": 1.3839,
+      "step": 6228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9136388301849365,
+      "learning_rate": 4.835606661379857e-05,
+      "loss": 1.463,
+      "step": 6229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8412753343582153,
+      "learning_rate": 4.8355802273328046e-05,
+      "loss": 1.5452,
+      "step": 6230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5588231086730957,
+      "learning_rate": 4.8355537932857527e-05,
+      "loss": 1.4515,
+      "step": 6231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0114216804504395,
+      "learning_rate": 4.835527359238699e-05,
+      "loss": 1.3525,
+      "step": 6232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7428319454193115,
+      "learning_rate": 4.8355009251916474e-05,
+      "loss": 1.2873,
+      "step": 6233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6115347146987915,
+      "learning_rate": 4.835474491144594e-05,
+      "loss": 1.4657,
+      "step": 6234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.748363971710205,
+      "learning_rate": 4.835448057097542e-05,
+      "loss": 1.3759,
+      "step": 6235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8454314470291138,
+      "learning_rate": 4.8354216230504894e-05,
+      "loss": 1.4991,
+      "step": 6236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9295296669006348,
+      "learning_rate": 4.835395189003437e-05,
+      "loss": 1.5718,
+      "step": 6237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.652147650718689,
+      "learning_rate": 4.835368754956384e-05,
+      "loss": 1.2569,
+      "step": 6238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.349867105484009,
+      "learning_rate": 4.8353423209093315e-05,
+      "loss": 1.5301,
+      "step": 6239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9153534173965454,
+      "learning_rate": 4.835315886862279e-05,
+      "loss": 1.5748,
+      "step": 6240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6923389434814453,
+      "learning_rate": 4.835289452815226e-05,
+      "loss": 1.4373,
+      "step": 6241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9573804140090942,
+      "learning_rate": 4.8352630187681736e-05,
+      "loss": 1.5174,
+      "step": 6242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5762145519256592,
+      "learning_rate": 4.835236584721121e-05,
+      "loss": 1.417,
+      "step": 6243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.821407675743103,
+      "learning_rate": 4.835210150674068e-05,
+      "loss": 1.4513,
+      "step": 6244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5011626482009888,
+      "learning_rate": 4.8351837166270157e-05,
+      "loss": 1.3454,
+      "step": 6245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9572265148162842,
+      "learning_rate": 4.835157282579964e-05,
+      "loss": 1.4016,
+      "step": 6246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9570163488388062,
+      "learning_rate": 4.8351308485329104e-05,
+      "loss": 1.608,
+      "step": 6247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.733722448348999,
+      "learning_rate": 4.8351044144858584e-05,
+      "loss": 1.4739,
+      "step": 6248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5413893461227417,
+      "learning_rate": 4.835077980438805e-05,
+      "loss": 1.4101,
+      "step": 6249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6841338872909546,
+      "learning_rate": 4.835051546391753e-05,
+      "loss": 1.4263,
+      "step": 6250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1462912559509277,
+      "learning_rate": 4.8350251123447005e-05,
+      "loss": 1.6186,
+      "step": 6251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0460293292999268,
+      "learning_rate": 4.834998678297648e-05,
+      "loss": 1.5663,
+      "step": 6252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7882461547851562,
+      "learning_rate": 4.834972244250595e-05,
+      "loss": 1.2987,
+      "step": 6253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4213255643844604,
+      "learning_rate": 4.834945810203542e-05,
+      "loss": 1.4063,
+      "step": 6254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5951813459396362,
+      "learning_rate": 4.83491937615649e-05,
+      "loss": 1.3654,
+      "step": 6255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.870391607284546,
+      "learning_rate": 4.834892942109437e-05,
+      "loss": 1.4612,
+      "step": 6256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6657761335372925,
+      "learning_rate": 4.8348665080623846e-05,
+      "loss": 1.3865,
+      "step": 6257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9230998754501343,
+      "learning_rate": 4.834840074015332e-05,
+      "loss": 1.5263,
+      "step": 6258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.421077847480774,
+      "learning_rate": 4.834813639968279e-05,
+      "loss": 1.2375,
+      "step": 6259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9061691761016846,
+      "learning_rate": 4.834787205921227e-05,
+      "loss": 1.4373,
+      "step": 6260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6494598388671875,
+      "learning_rate": 4.834760771874175e-05,
+      "loss": 1.3178,
+      "step": 6261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.799357295036316,
+      "learning_rate": 4.8347343378271214e-05,
+      "loss": 1.4862,
+      "step": 6262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0310275554656982,
+      "learning_rate": 4.8347079037800694e-05,
+      "loss": 1.6656,
+      "step": 6263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.631926417350769,
+      "learning_rate": 4.834681469733016e-05,
+      "loss": 1.2909,
+      "step": 6264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.053513765335083,
+      "learning_rate": 4.8346550356859635e-05,
+      "loss": 1.3098,
+      "step": 6265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.780081868171692,
+      "learning_rate": 4.8346286016389115e-05,
+      "loss": 1.25,
+      "step": 6266
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9026477336883545,
+      "learning_rate": 4.834602167591858e-05,
+      "loss": 1.5275,
+      "step": 6267
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.474252462387085,
+      "learning_rate": 4.834575733544806e-05,
+      "loss": 1.5606,
+      "step": 6268
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.04929256439209,
+      "learning_rate": 4.834549299497753e-05,
+      "loss": 1.7169,
+      "step": 6269
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.551708698272705,
+      "learning_rate": 4.834522865450701e-05,
+      "loss": 1.3804,
+      "step": 6270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5765008926391602,
+      "learning_rate": 4.834496431403648e-05,
+      "loss": 1.3957,
+      "step": 6271
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7408407926559448,
+      "learning_rate": 4.8344699973565957e-05,
+      "loss": 1.2538,
+      "step": 6272
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.077242612838745,
+      "learning_rate": 4.834443563309543e-05,
+      "loss": 1.4645,
+      "step": 6273
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0155699253082275,
+      "learning_rate": 4.8344171292624904e-05,
+      "loss": 1.5452,
+      "step": 6274
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.676321268081665,
+      "learning_rate": 4.834390695215438e-05,
+      "loss": 1.492,
+      "step": 6275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.330888032913208,
+      "learning_rate": 4.834364261168385e-05,
+      "loss": 1.5154,
+      "step": 6276
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.94794762134552,
+      "learning_rate": 4.8343378271213324e-05,
+      "loss": 1.4468,
+      "step": 6277
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7204158306121826,
+      "learning_rate": 4.83431139307428e-05,
+      "loss": 1.4236,
+      "step": 6278
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8757145404815674,
+      "learning_rate": 4.834284959027227e-05,
+      "loss": 1.3294,
+      "step": 6279
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.727523922920227,
+      "learning_rate": 4.8342585249801745e-05,
+      "loss": 1.1852,
+      "step": 6280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1370608806610107,
+      "learning_rate": 4.8342320909331225e-05,
+      "loss": 1.6931,
+      "step": 6281
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0934548377990723,
+      "learning_rate": 4.834205656886069e-05,
+      "loss": 1.3917,
+      "step": 6282
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3213063478469849,
+      "learning_rate": 4.834179222839017e-05,
+      "loss": 1.2279,
+      "step": 6283
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.435080647468567,
+      "learning_rate": 4.834152788791964e-05,
+      "loss": 1.4395,
+      "step": 6284
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.982160210609436,
+      "learning_rate": 4.834126354744912e-05,
+      "loss": 1.4399,
+      "step": 6285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.69691002368927,
+      "learning_rate": 4.834099920697859e-05,
+      "loss": 1.4063,
+      "step": 6286
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4004462957382202,
+      "learning_rate": 4.834073486650807e-05,
+      "loss": 1.384,
+      "step": 6287
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6872142553329468,
+      "learning_rate": 4.834047052603754e-05,
+      "loss": 1.7364,
+      "step": 6288
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4095920324325562,
+      "learning_rate": 4.8340206185567014e-05,
+      "loss": 1.2525,
+      "step": 6289
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.58470618724823,
+      "learning_rate": 4.833994184509649e-05,
+      "loss": 1.4415,
+      "step": 6290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8406845331192017,
+      "learning_rate": 4.833967750462596e-05,
+      "loss": 1.358,
+      "step": 6291
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6405870914459229,
+      "learning_rate": 4.8339413164155435e-05,
+      "loss": 1.2564,
+      "step": 6292
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8240808248519897,
+      "learning_rate": 4.833914882368491e-05,
+      "loss": 1.3489,
+      "step": 6293
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6139470338821411,
+      "learning_rate": 4.833888448321438e-05,
+      "loss": 1.2635,
+      "step": 6294
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9618715047836304,
+      "learning_rate": 4.8338620142743855e-05,
+      "loss": 1.4917,
+      "step": 6295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.16302227973938,
+      "learning_rate": 4.8338355802273336e-05,
+      "loss": 1.4266,
+      "step": 6296
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0831797122955322,
+      "learning_rate": 4.83380914618028e-05,
+      "loss": 1.4557,
+      "step": 6297
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9785611629486084,
+      "learning_rate": 4.833782712133228e-05,
+      "loss": 1.4279,
+      "step": 6298
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6151162385940552,
+      "learning_rate": 4.833756278086175e-05,
+      "loss": 1.3673,
+      "step": 6299
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4005463123321533,
+      "learning_rate": 4.833729844039122e-05,
+      "loss": 1.2076,
+      "step": 6300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.119279146194458,
+      "learning_rate": 4.8337034099920704e-05,
+      "loss": 1.5694,
+      "step": 6301
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2468223571777344,
+      "learning_rate": 4.833676975945017e-05,
+      "loss": 1.5616,
+      "step": 6302
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.023087978363037,
+      "learning_rate": 4.833650541897965e-05,
+      "loss": 1.4024,
+      "step": 6303
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.459419846534729,
+      "learning_rate": 4.833624107850912e-05,
+      "loss": 1.2307,
+      "step": 6304
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4514317512512207,
+      "learning_rate": 4.83359767380386e-05,
+      "loss": 1.3641,
+      "step": 6305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4771945476531982,
+      "learning_rate": 4.833571239756807e-05,
+      "loss": 1.4867,
+      "step": 6306
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.552785873413086,
+      "learning_rate": 4.8335448057097545e-05,
+      "loss": 1.6294,
+      "step": 6307
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5132993459701538,
+      "learning_rate": 4.833518371662702e-05,
+      "loss": 1.383,
+      "step": 6308
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.286914348602295,
+      "learning_rate": 4.833491937615649e-05,
+      "loss": 1.6006,
+      "step": 6309
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6452833414077759,
+      "learning_rate": 4.8334655035685966e-05,
+      "loss": 1.3825,
+      "step": 6310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4292324781417847,
+      "learning_rate": 4.833439069521544e-05,
+      "loss": 1.2844,
+      "step": 6311
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.568099856376648,
+      "learning_rate": 4.833412635474491e-05,
+      "loss": 1.227,
+      "step": 6312
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.501970887184143,
+      "learning_rate": 4.8333862014274386e-05,
+      "loss": 1.304,
+      "step": 6313
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5621672868728638,
+      "learning_rate": 4.833359767380386e-05,
+      "loss": 1.3061,
+      "step": 6314
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.211329460144043,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 1.4769,
+      "step": 6315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6707063913345337,
+      "learning_rate": 4.8333068992862814e-05,
+      "loss": 1.4077,
+      "step": 6316
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.66978919506073,
+      "learning_rate": 4.833280465239228e-05,
+      "loss": 1.4334,
+      "step": 6317
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5584771633148193,
+      "learning_rate": 4.833254031192176e-05,
+      "loss": 1.4702,
+      "step": 6318
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.936958909034729,
+      "learning_rate": 4.833227597145123e-05,
+      "loss": 1.4249,
+      "step": 6319
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7405692338943481,
+      "learning_rate": 4.833201163098071e-05,
+      "loss": 1.3351,
+      "step": 6320
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.792340636253357,
+      "learning_rate": 4.833174729051018e-05,
+      "loss": 1.3605,
+      "step": 6321
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9662531614303589,
+      "learning_rate": 4.8331482950039655e-05,
+      "loss": 1.3877,
+      "step": 6322
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.13358736038208,
+      "learning_rate": 4.833121860956913e-05,
+      "loss": 1.3111,
+      "step": 6323
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1035406589508057,
+      "learning_rate": 4.83309542690986e-05,
+      "loss": 1.589,
+      "step": 6324
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1719212532043457,
+      "learning_rate": 4.8330689928628076e-05,
+      "loss": 1.4504,
+      "step": 6325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7191102504730225,
+      "learning_rate": 4.833042558815755e-05,
+      "loss": 1.4227,
+      "step": 6326
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.898293137550354,
+      "learning_rate": 4.833016124768702e-05,
+      "loss": 1.3027,
+      "step": 6327
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7856141328811646,
+      "learning_rate": 4.83298969072165e-05,
+      "loss": 1.4477,
+      "step": 6328
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4620052576065063,
+      "learning_rate": 4.832963256674597e-05,
+      "loss": 1.3783,
+      "step": 6329
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9467923641204834,
+      "learning_rate": 4.8329368226275444e-05,
+      "loss": 1.4743,
+      "step": 6330
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4393846988677979,
+      "learning_rate": 4.8329103885804924e-05,
+      "loss": 1.3078,
+      "step": 6331
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.270690679550171,
+      "learning_rate": 4.832883954533439e-05,
+      "loss": 1.2334,
+      "step": 6332
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.616904377937317,
+      "learning_rate": 4.832857520486387e-05,
+      "loss": 1.4011,
+      "step": 6333
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.935329794883728,
+      "learning_rate": 4.832831086439334e-05,
+      "loss": 1.4648,
+      "step": 6334
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6650381088256836,
+      "learning_rate": 4.832804652392282e-05,
+      "loss": 1.2608,
+      "step": 6335
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9446316957473755,
+      "learning_rate": 4.832778218345229e-05,
+      "loss": 2.1025,
+      "step": 6336
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.081575393676758,
+      "learning_rate": 4.832751784298176e-05,
+      "loss": 1.4474,
+      "step": 6337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7519952058792114,
+      "learning_rate": 4.832725350251124e-05,
+      "loss": 1.3695,
+      "step": 6338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8710861206054688,
+      "learning_rate": 4.8326989162040706e-05,
+      "loss": 1.4623,
+      "step": 6339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.945740818977356,
+      "learning_rate": 4.8326724821570186e-05,
+      "loss": 1.3502,
+      "step": 6340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6733973026275635,
+      "learning_rate": 4.832646048109966e-05,
+      "loss": 1.4971,
+      "step": 6341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.582993507385254,
+      "learning_rate": 4.8326196140629134e-05,
+      "loss": 1.2486,
+      "step": 6342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.837533473968506,
+      "learning_rate": 4.832593180015861e-05,
+      "loss": 1.3099,
+      "step": 6343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9119889736175537,
+      "learning_rate": 4.832566745968808e-05,
+      "loss": 1.3324,
+      "step": 6344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2908813953399658,
+      "learning_rate": 4.8325403119217554e-05,
+      "loss": 1.2106,
+      "step": 6345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4370369911193848,
+      "learning_rate": 4.832513877874703e-05,
+      "loss": 1.4251,
+      "step": 6346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6502221822738647,
+      "learning_rate": 4.83248744382765e-05,
+      "loss": 1.2548,
+      "step": 6347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5441622734069824,
+      "learning_rate": 4.8324610097805975e-05,
+      "loss": 1.5795,
+      "step": 6348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6888738870620728,
+      "learning_rate": 4.832434575733545e-05,
+      "loss": 1.4445,
+      "step": 6349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.24250864982605,
+      "learning_rate": 4.832408141686492e-05,
+      "loss": 1.5099,
+      "step": 6350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6678537130355835,
+      "learning_rate": 4.83238170763944e-05,
+      "loss": 1.3868,
+      "step": 6351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8835961818695068,
+      "learning_rate": 4.832355273592387e-05,
+      "loss": 1.3841,
+      "step": 6352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6475833654403687,
+      "learning_rate": 4.832328839545335e-05,
+      "loss": 1.421,
+      "step": 6353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8876315355300903,
+      "learning_rate": 4.8323024054982816e-05,
+      "loss": 1.4666,
+      "step": 6354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.239565372467041,
+      "learning_rate": 4.83227597145123e-05,
+      "loss": 1.4647,
+      "step": 6355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7688617706298828,
+      "learning_rate": 4.832249537404177e-05,
+      "loss": 1.4819,
+      "step": 6356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5883604288101196,
+      "learning_rate": 4.8322231033571244e-05,
+      "loss": 1.3211,
+      "step": 6357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9604363441467285,
+      "learning_rate": 4.832196669310072e-05,
+      "loss": 1.3165,
+      "step": 6358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9577630758285522,
+      "learning_rate": 4.832170235263019e-05,
+      "loss": 1.3217,
+      "step": 6359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0031721591949463,
+      "learning_rate": 4.8321438012159665e-05,
+      "loss": 1.3171,
+      "step": 6360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5793020725250244,
+      "learning_rate": 4.832117367168914e-05,
+      "loss": 1.3392,
+      "step": 6361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8520294427871704,
+      "learning_rate": 4.832090933121861e-05,
+      "loss": 1.3876,
+      "step": 6362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2494285106658936,
+      "learning_rate": 4.8320644990748085e-05,
+      "loss": 1.5101,
+      "step": 6363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1278233528137207,
+      "learning_rate": 4.832038065027756e-05,
+      "loss": 1.3164,
+      "step": 6364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.619181513786316,
+      "learning_rate": 4.832011630980703e-05,
+      "loss": 1.5637,
+      "step": 6365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5873709917068481,
+      "learning_rate": 4.831985196933651e-05,
+      "loss": 1.3109,
+      "step": 6366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6933521032333374,
+      "learning_rate": 4.831958762886598e-05,
+      "loss": 1.2977,
+      "step": 6367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6637964248657227,
+      "learning_rate": 4.831932328839546e-05,
+      "loss": 1.1816,
+      "step": 6368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.228384017944336,
+      "learning_rate": 4.831905894792493e-05,
+      "loss": 1.5049,
+      "step": 6369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6777584552764893,
+      "learning_rate": 4.831879460745441e-05,
+      "loss": 1.7516,
+      "step": 6370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.650399684906006,
+      "learning_rate": 4.831853026698388e-05,
+      "loss": 1.5969,
+      "step": 6371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5617526769638062,
+      "learning_rate": 4.831826592651335e-05,
+      "loss": 1.3369,
+      "step": 6372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9534616470336914,
+      "learning_rate": 4.831800158604283e-05,
+      "loss": 1.4771,
+      "step": 6373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7180496454238892,
+      "learning_rate": 4.8317737245572295e-05,
+      "loss": 1.3994,
+      "step": 6374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1593031883239746,
+      "learning_rate": 4.8317472905101775e-05,
+      "loss": 1.3997,
+      "step": 6375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.500563383102417,
+      "learning_rate": 4.831720856463125e-05,
+      "loss": 1.3454,
+      "step": 6376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6952048540115356,
+      "learning_rate": 4.831694422416072e-05,
+      "loss": 1.3917,
+      "step": 6377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7531423568725586,
+      "learning_rate": 4.8316679883690196e-05,
+      "loss": 1.4655,
+      "step": 6378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.593546986579895,
+      "learning_rate": 4.831641554321967e-05,
+      "loss": 1.303,
+      "step": 6379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6888447999954224,
+      "learning_rate": 4.831615120274914e-05,
+      "loss": 1.4187,
+      "step": 6380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8994653224945068,
+      "learning_rate": 4.831588686227862e-05,
+      "loss": 1.5081,
+      "step": 6381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4880824089050293,
+      "learning_rate": 4.831562252180809e-05,
+      "loss": 1.3744,
+      "step": 6382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6884756088256836,
+      "learning_rate": 4.8315358181337564e-05,
+      "loss": 1.8075,
+      "step": 6383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0865209102630615,
+      "learning_rate": 4.831509384086704e-05,
+      "loss": 1.5487,
+      "step": 6384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.12733268737793,
+      "learning_rate": 4.831482950039651e-05,
+      "loss": 1.6119,
+      "step": 6385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.221874952316284,
+      "learning_rate": 4.831456515992599e-05,
+      "loss": 2.6141,
+      "step": 6386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3285577297210693,
+      "learning_rate": 4.831430081945546e-05,
+      "loss": 1.3769,
+      "step": 6387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7611072063446045,
+      "learning_rate": 4.831403647898494e-05,
+      "loss": 1.2373,
+      "step": 6388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6899112462997437,
+      "learning_rate": 4.8313772138514405e-05,
+      "loss": 1.3612,
+      "step": 6389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6497329473495483,
+      "learning_rate": 4.8313507798043885e-05,
+      "loss": 1.4138,
+      "step": 6390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6913704872131348,
+      "learning_rate": 4.831324345757336e-05,
+      "loss": 1.5081,
+      "step": 6391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0003244876861572,
+      "learning_rate": 4.831297911710283e-05,
+      "loss": 1.5649,
+      "step": 6392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5064146518707275,
+      "learning_rate": 4.8312714776632306e-05,
+      "loss": 1.2717,
+      "step": 6393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.64133882522583,
+      "learning_rate": 4.831245043616178e-05,
+      "loss": 1.4435,
+      "step": 6394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.514545202255249,
+      "learning_rate": 4.831218609569125e-05,
+      "loss": 1.6141,
+      "step": 6395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8108103275299072,
+      "learning_rate": 4.831192175522073e-05,
+      "loss": 1.231,
+      "step": 6396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5008295774459839,
+      "learning_rate": 4.83116574147502e-05,
+      "loss": 1.4323,
+      "step": 6397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.052105665206909,
+      "learning_rate": 4.8311393074279674e-05,
+      "loss": 1.425,
+      "step": 6398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7254438400268555,
+      "learning_rate": 4.831112873380915e-05,
+      "loss": 1.2369,
+      "step": 6399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1315596103668213,
+      "learning_rate": 4.831086439333862e-05,
+      "loss": 1.5192,
+      "step": 6400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5799283981323242,
+      "learning_rate": 4.83106000528681e-05,
+      "loss": 1.4082,
+      "step": 6401
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.287457227706909,
+      "learning_rate": 4.831033571239757e-05,
+      "loss": 1.3789,
+      "step": 6402
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6936540603637695,
+      "learning_rate": 4.831007137192705e-05,
+      "loss": 1.8533,
+      "step": 6403
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8773046731948853,
+      "learning_rate": 4.8309807031456515e-05,
+      "loss": 1.4665,
+      "step": 6404
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6478769779205322,
+      "learning_rate": 4.8309542690985996e-05,
+      "loss": 1.4165,
+      "step": 6405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8928309679031372,
+      "learning_rate": 4.830927835051547e-05,
+      "loss": 1.379,
+      "step": 6406
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.513704776763916,
+      "learning_rate": 4.830901401004494e-05,
+      "loss": 1.4465,
+      "step": 6407
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4676852226257324,
+      "learning_rate": 4.8308749669574416e-05,
+      "loss": 1.3027,
+      "step": 6408
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3673810958862305,
+      "learning_rate": 4.830848532910388e-05,
+      "loss": 1.4472,
+      "step": 6409
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0571448802948,
+      "learning_rate": 4.8308220988633364e-05,
+      "loss": 1.2366,
+      "step": 6410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.637710690498352,
+      "learning_rate": 4.830795664816284e-05,
+      "loss": 1.5401,
+      "step": 6411
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.425175428390503,
+      "learning_rate": 4.830769230769231e-05,
+      "loss": 1.1975,
+      "step": 6412
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.708994746208191,
+      "learning_rate": 4.8307427967221784e-05,
+      "loss": 1.4877,
+      "step": 6413
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4024239778518677,
+      "learning_rate": 4.830716362675126e-05,
+      "loss": 1.3613,
+      "step": 6414
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5860768556594849,
+      "learning_rate": 4.830689928628073e-05,
+      "loss": 1.4667,
+      "step": 6415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5692943334579468,
+      "learning_rate": 4.830663494581021e-05,
+      "loss": 1.3357,
+      "step": 6416
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.213101387023926,
+      "learning_rate": 4.830637060533968e-05,
+      "loss": 1.5304,
+      "step": 6417
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5646586418151855,
+      "learning_rate": 4.830610626486915e-05,
+      "loss": 1.4222,
+      "step": 6418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.595931053161621,
+      "learning_rate": 4.8305841924398626e-05,
+      "loss": 1.3673,
+      "step": 6419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5690195560455322,
+      "learning_rate": 4.83055775839281e-05,
+      "loss": 1.2424,
+      "step": 6420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6400318145751953,
+      "learning_rate": 4.830531324345758e-05,
+      "loss": 1.8818,
+      "step": 6421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8822014331817627,
+      "learning_rate": 4.8305048902987046e-05,
+      "loss": 1.3861,
+      "step": 6422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.561154842376709,
+      "learning_rate": 4.830478456251653e-05,
+      "loss": 1.5306,
+      "step": 6423
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4670372009277344,
+      "learning_rate": 4.8304520222045994e-05,
+      "loss": 1.3458,
+      "step": 6424
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7730599641799927,
+      "learning_rate": 4.8304255881575474e-05,
+      "loss": 1.319,
+      "step": 6425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4865262508392334,
+      "learning_rate": 4.830399154110495e-05,
+      "loss": 1.3755,
+      "step": 6426
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0380873680114746,
+      "learning_rate": 4.830372720063442e-05,
+      "loss": 1.4316,
+      "step": 6427
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5877580642700195,
+      "learning_rate": 4.8303462860163895e-05,
+      "loss": 1.4769,
+      "step": 6428
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6067368984222412,
+      "learning_rate": 4.830319851969337e-05,
+      "loss": 1.3064,
+      "step": 6429
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6847563982009888,
+      "learning_rate": 4.830293417922284e-05,
+      "loss": 2.0264,
+      "step": 6430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.702920436859131,
+      "learning_rate": 4.8302669838752315e-05,
+      "loss": 1.4164,
+      "step": 6431
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6627713441848755,
+      "learning_rate": 4.830240549828179e-05,
+      "loss": 1.4176,
+      "step": 6432
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.615374803543091,
+      "learning_rate": 4.830214115781126e-05,
+      "loss": 1.5705,
+      "step": 6433
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5711934566497803,
+      "learning_rate": 4.8301876817340736e-05,
+      "loss": 1.2941,
+      "step": 6434
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6282368898391724,
+      "learning_rate": 4.830161247687021e-05,
+      "loss": 1.4705,
+      "step": 6435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0929033756256104,
+      "learning_rate": 4.830134813639969e-05,
+      "loss": 1.4955,
+      "step": 6436
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.249404191970825,
+      "learning_rate": 4.830108379592916e-05,
+      "loss": 1.1981,
+      "step": 6437
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.046019554138184,
+      "learning_rate": 4.830081945545864e-05,
+      "loss": 1.5993,
+      "step": 6438
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0686681270599365,
+      "learning_rate": 4.8300555114988104e-05,
+      "loss": 1.5472,
+      "step": 6439
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5413074493408203,
+      "learning_rate": 4.8300290774517584e-05,
+      "loss": 1.349,
+      "step": 6440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.737112283706665,
+      "learning_rate": 4.830002643404706e-05,
+      "loss": 1.3458,
+      "step": 6441
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6027660369873047,
+      "learning_rate": 4.829976209357653e-05,
+      "loss": 1.34,
+      "step": 6442
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4129754304885864,
+      "learning_rate": 4.8299497753106005e-05,
+      "loss": 1.4756,
+      "step": 6443
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.671401858329773,
+      "learning_rate": 4.829923341263547e-05,
+      "loss": 1.3377,
+      "step": 6444
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.162015914916992,
+      "learning_rate": 4.829896907216495e-05,
+      "loss": 1.4461,
+      "step": 6445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9755018949508667,
+      "learning_rate": 4.8298704731694426e-05,
+      "loss": 1.4057,
+      "step": 6446
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.124779224395752,
+      "learning_rate": 4.82984403912239e-05,
+      "loss": 1.3952,
+      "step": 6447
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8724548816680908,
+      "learning_rate": 4.829817605075337e-05,
+      "loss": 1.4441,
+      "step": 6448
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7387179136276245,
+      "learning_rate": 4.8297911710282846e-05,
+      "loss": 1.4829,
+      "step": 6449
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8525272607803345,
+      "learning_rate": 4.829764736981232e-05,
+      "loss": 1.3746,
+      "step": 6450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9835914373397827,
+      "learning_rate": 4.82973830293418e-05,
+      "loss": 1.6115,
+      "step": 6451
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9894511699676514,
+      "learning_rate": 4.829711868887127e-05,
+      "loss": 1.579,
+      "step": 6452
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6901943683624268,
+      "learning_rate": 4.829685434840075e-05,
+      "loss": 1.4041,
+      "step": 6453
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6186683177948,
+      "learning_rate": 4.8296590007930214e-05,
+      "loss": 1.5961,
+      "step": 6454
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5841108560562134,
+      "learning_rate": 4.829632566745969e-05,
+      "loss": 1.3562,
+      "step": 6455
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8112678527832031,
+      "learning_rate": 4.829606132698917e-05,
+      "loss": 1.4185,
+      "step": 6456
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0523734092712402,
+      "learning_rate": 4.8295796986518635e-05,
+      "loss": 1.5814,
+      "step": 6457
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7291287183761597,
+      "learning_rate": 4.8295532646048115e-05,
+      "loss": 1.2781,
+      "step": 6458
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6482768058776855,
+      "learning_rate": 4.829526830557758e-05,
+      "loss": 1.4395,
+      "step": 6459
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.521763563156128,
+      "learning_rate": 4.829500396510706e-05,
+      "loss": 1.418,
+      "step": 6460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.526678442955017,
+      "learning_rate": 4.8294739624636536e-05,
+      "loss": 1.305,
+      "step": 6461
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.902441143989563,
+      "learning_rate": 4.829447528416601e-05,
+      "loss": 1.4322,
+      "step": 6462
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.982157826423645,
+      "learning_rate": 4.829421094369548e-05,
+      "loss": 1.4213,
+      "step": 6463
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7168077230453491,
+      "learning_rate": 4.829394660322496e-05,
+      "loss": 1.4987,
+      "step": 6464
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.729477882385254,
+      "learning_rate": 4.829368226275443e-05,
+      "loss": 1.4236,
+      "step": 6465
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5829261541366577,
+      "learning_rate": 4.8293417922283904e-05,
+      "loss": 1.3188,
+      "step": 6466
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.979852318763733,
+      "learning_rate": 4.829315358181338e-05,
+      "loss": 1.488,
+      "step": 6467
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3884581327438354,
+      "learning_rate": 4.829288924134285e-05,
+      "loss": 1.2574,
+      "step": 6468
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4800852537155151,
+      "learning_rate": 4.8292624900872325e-05,
+      "loss": 1.2315,
+      "step": 6469
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.741776466369629,
+      "learning_rate": 4.82923605604018e-05,
+      "loss": 1.3834,
+      "step": 6470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6861647367477417,
+      "learning_rate": 4.829209621993128e-05,
+      "loss": 1.3268,
+      "step": 6471
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8613049983978271,
+      "learning_rate": 4.8291831879460745e-05,
+      "loss": 1.4199,
+      "step": 6472
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8228870630264282,
+      "learning_rate": 4.8291567538990226e-05,
+      "loss": 1.4352,
+      "step": 6473
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3924598693847656,
+      "learning_rate": 4.829130319851969e-05,
+      "loss": 1.6226,
+      "step": 6474
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3296854496002197,
+      "learning_rate": 4.829103885804917e-05,
+      "loss": 1.2595,
+      "step": 6475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0309031009674072,
+      "learning_rate": 4.8290774517578646e-05,
+      "loss": 1.8183,
+      "step": 6476
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1441283226013184,
+      "learning_rate": 4.829051017710812e-05,
+      "loss": 1.47,
+      "step": 6477
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6385304927825928,
+      "learning_rate": 4.8290245836637593e-05,
+      "loss": 1.2076,
+      "step": 6478
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8669763803482056,
+      "learning_rate": 4.828998149616707e-05,
+      "loss": 1.5587,
+      "step": 6479
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.33193039894104,
+      "learning_rate": 4.828971715569654e-05,
+      "loss": 1.2351,
+      "step": 6480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.468520164489746,
+      "learning_rate": 4.8289452815226014e-05,
+      "loss": 1.2985,
+      "step": 6481
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6480127573013306,
+      "learning_rate": 4.828918847475549e-05,
+      "loss": 1.4275,
+      "step": 6482
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4805864095687866,
+      "learning_rate": 4.828892413428496e-05,
+      "loss": 1.3657,
+      "step": 6483
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.665002703666687,
+      "learning_rate": 4.8288659793814435e-05,
+      "loss": 1.485,
+      "step": 6484
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.303785562515259,
+      "learning_rate": 4.828839545334391e-05,
+      "loss": 1.4416,
+      "step": 6485
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.708299994468689,
+      "learning_rate": 4.828813111287339e-05,
+      "loss": 1.2793,
+      "step": 6486
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.269939661026001,
+      "learning_rate": 4.8287866772402856e-05,
+      "loss": 1.4901,
+      "step": 6487
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.864259958267212,
+      "learning_rate": 4.8287602431932336e-05,
+      "loss": 1.4703,
+      "step": 6488
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8140348196029663,
+      "learning_rate": 4.82873380914618e-05,
+      "loss": 1.2764,
+      "step": 6489
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8246734142303467,
+      "learning_rate": 4.8287073750991276e-05,
+      "loss": 1.3834,
+      "step": 6490
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7758578062057495,
+      "learning_rate": 4.828680941052076e-05,
+      "loss": 1.3669,
+      "step": 6491
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7276962995529175,
+      "learning_rate": 4.8286545070050223e-05,
+      "loss": 1.3782,
+      "step": 6492
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7210311889648438,
+      "learning_rate": 4.8286280729579704e-05,
+      "loss": 1.3933,
+      "step": 6493
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4435694217681885,
+      "learning_rate": 4.828601638910917e-05,
+      "loss": 1.6057,
+      "step": 6494
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8763900995254517,
+      "learning_rate": 4.828575204863865e-05,
+      "loss": 1.4932,
+      "step": 6495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4409087896347046,
+      "learning_rate": 4.8285487708168124e-05,
+      "loss": 1.3153,
+      "step": 6496
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8426872491836548,
+      "learning_rate": 4.82852233676976e-05,
+      "loss": 1.5092,
+      "step": 6497
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6939337253570557,
+      "learning_rate": 4.828495902722707e-05,
+      "loss": 1.4222,
+      "step": 6498
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2502248287200928,
+      "learning_rate": 4.8284694686756545e-05,
+      "loss": 1.412,
+      "step": 6499
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.267145156860352,
+      "learning_rate": 4.828443034628602e-05,
+      "loss": 1.4773,
+      "step": 6500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.481999158859253,
+      "learning_rate": 4.828416600581549e-05,
+      "loss": 1.6289,
+      "step": 6501
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7537519931793213,
+      "learning_rate": 4.8283901665344966e-05,
+      "loss": 1.0996,
+      "step": 6502
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.58809494972229,
+      "learning_rate": 4.828363732487444e-05,
+      "loss": 1.3483,
+      "step": 6503
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7819209098815918,
+      "learning_rate": 4.828337298440391e-05,
+      "loss": 1.1652,
+      "step": 6504
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6174354553222656,
+      "learning_rate": 4.828310864393339e-05,
+      "loss": 1.3533,
+      "step": 6505
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9028514623641968,
+      "learning_rate": 4.828284430346287e-05,
+      "loss": 1.3035,
+      "step": 6506
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8383848667144775,
+      "learning_rate": 4.8282579962992334e-05,
+      "loss": 1.476,
+      "step": 6507
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8484656810760498,
+      "learning_rate": 4.8282315622521814e-05,
+      "loss": 1.408,
+      "step": 6508
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.209935426712036,
+      "learning_rate": 4.828205128205128e-05,
+      "loss": 1.3865,
+      "step": 6509
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3629523515701294,
+      "learning_rate": 4.828178694158076e-05,
+      "loss": 1.3597,
+      "step": 6510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4083796739578247,
+      "learning_rate": 4.8281522601110235e-05,
+      "loss": 1.3467,
+      "step": 6511
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.508018732070923,
+      "learning_rate": 4.828125826063971e-05,
+      "loss": 1.4644,
+      "step": 6512
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7603119611740112,
+      "learning_rate": 4.828099392016918e-05,
+      "loss": 1.4642,
+      "step": 6513
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6528749465942383,
+      "learning_rate": 4.8280729579698656e-05,
+      "loss": 1.4628,
+      "step": 6514
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8818676471710205,
+      "learning_rate": 4.828046523922813e-05,
+      "loss": 1.4086,
+      "step": 6515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5516552925109863,
+      "learning_rate": 4.82802008987576e-05,
+      "loss": 1.483,
+      "step": 6516
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3695719242095947,
+      "learning_rate": 4.8279936558287076e-05,
+      "loss": 1.5549,
+      "step": 6517
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.501499891281128,
+      "learning_rate": 4.827967221781655e-05,
+      "loss": 1.2872,
+      "step": 6518
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5501999855041504,
+      "learning_rate": 4.8279407877346023e-05,
+      "loss": 1.4144,
+      "step": 6519
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7774828672409058,
+      "learning_rate": 4.82791435368755e-05,
+      "loss": 1.3931,
+      "step": 6520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5547235012054443,
+      "learning_rate": 4.827887919640497e-05,
+      "loss": 1.2438,
+      "step": 6521
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.737828254699707,
+      "learning_rate": 4.8278614855934444e-05,
+      "loss": 1.8252,
+      "step": 6522
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.536944031715393,
+      "learning_rate": 4.8278350515463924e-05,
+      "loss": 1.5586,
+      "step": 6523
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.610807180404663,
+      "learning_rate": 4.827808617499339e-05,
+      "loss": 1.4875,
+      "step": 6524
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9317232370376587,
+      "learning_rate": 4.827782183452287e-05,
+      "loss": 1.4972,
+      "step": 6525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9802590608596802,
+      "learning_rate": 4.827755749405234e-05,
+      "loss": 1.476,
+      "step": 6526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4351516962051392,
+      "learning_rate": 4.827729315358181e-05,
+      "loss": 1.2884,
+      "step": 6527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3406383991241455,
+      "learning_rate": 4.827702881311129e-05,
+      "loss": 1.5058,
+      "step": 6528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8911575078964233,
+      "learning_rate": 4.827676447264076e-05,
+      "loss": 1.374,
+      "step": 6529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4907433986663818,
+      "learning_rate": 4.827650013217024e-05,
+      "loss": 1.4579,
+      "step": 6530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3951973915100098,
+      "learning_rate": 4.8276235791699706e-05,
+      "loss": 1.6579,
+      "step": 6531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2652339935302734,
+      "learning_rate": 4.8275971451229187e-05,
+      "loss": 1.4734,
+      "step": 6532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5338433980941772,
+      "learning_rate": 4.827570711075866e-05,
+      "loss": 1.2795,
+      "step": 6533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1734845638275146,
+      "learning_rate": 4.8275442770288134e-05,
+      "loss": 1.4696,
+      "step": 6534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7216604948043823,
+      "learning_rate": 4.827517842981761e-05,
+      "loss": 1.5318,
+      "step": 6535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2089349031448364,
+      "learning_rate": 4.827491408934708e-05,
+      "loss": 1.4013,
+      "step": 6536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8604034185409546,
+      "learning_rate": 4.8274649748876554e-05,
+      "loss": 1.3754,
+      "step": 6537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.587526798248291,
+      "learning_rate": 4.827438540840603e-05,
+      "loss": 1.4089,
+      "step": 6538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6591250896453857,
+      "learning_rate": 4.82741210679355e-05,
+      "loss": 1.4736,
+      "step": 6539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.358572006225586,
+      "learning_rate": 4.8273856727464975e-05,
+      "loss": 1.3346,
+      "step": 6540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7856941223144531,
+      "learning_rate": 4.827359238699445e-05,
+      "loss": 1.5289,
+      "step": 6541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5515588521957397,
+      "learning_rate": 4.827332804652392e-05,
+      "loss": 1.4058,
+      "step": 6542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.732752799987793,
+      "learning_rate": 4.82730637060534e-05,
+      "loss": 1.4717,
+      "step": 6543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6413851976394653,
+      "learning_rate": 4.827279936558287e-05,
+      "loss": 1.3864,
+      "step": 6544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9391218423843384,
+      "learning_rate": 4.827253502511235e-05,
+      "loss": 1.6175,
+      "step": 6545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8209447860717773,
+      "learning_rate": 4.8272270684641817e-05,
+      "loss": 1.4874,
+      "step": 6546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.585849404335022,
+      "learning_rate": 4.82720063441713e-05,
+      "loss": 1.2396,
+      "step": 6547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.29647696018219,
+      "learning_rate": 4.827174200370077e-05,
+      "loss": 1.1801,
+      "step": 6548
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9658786058425903,
+      "learning_rate": 4.8271477663230244e-05,
+      "loss": 1.5638,
+      "step": 6549
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9680671691894531,
+      "learning_rate": 4.827121332275972e-05,
+      "loss": 1.2344,
+      "step": 6550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.403441309928894,
+      "learning_rate": 4.827094898228919e-05,
+      "loss": 1.2132,
+      "step": 6551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6186281442642212,
+      "learning_rate": 4.8270684641818665e-05,
+      "loss": 1.3087,
+      "step": 6552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7063770294189453,
+      "learning_rate": 4.827042030134814e-05,
+      "loss": 1.4769,
+      "step": 6553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.852573275566101,
+      "learning_rate": 4.827015596087761e-05,
+      "loss": 1.2973,
+      "step": 6554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6482906341552734,
+      "learning_rate": 4.8269891620407086e-05,
+      "loss": 1.6341,
+      "step": 6555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7544453144073486,
+      "learning_rate": 4.826962727993656e-05,
+      "loss": 1.5248,
+      "step": 6556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6569429636001587,
+      "learning_rate": 4.826936293946603e-05,
+      "loss": 1.5415,
+      "step": 6557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.377422571182251,
+      "learning_rate": 4.826909859899551e-05,
+      "loss": 1.3988,
+      "step": 6558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9277148246765137,
+      "learning_rate": 4.826883425852498e-05,
+      "loss": 1.4797,
+      "step": 6559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.852426528930664,
+      "learning_rate": 4.826856991805446e-05,
+      "loss": 1.3862,
+      "step": 6560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6471315622329712,
+      "learning_rate": 4.826830557758393e-05,
+      "loss": 1.6019,
+      "step": 6561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7324570417404175,
+      "learning_rate": 4.82680412371134e-05,
+      "loss": 2.1295,
+      "step": 6562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.393152713775635,
+      "learning_rate": 4.826777689664288e-05,
+      "loss": 1.5094,
+      "step": 6563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7544339895248413,
+      "learning_rate": 4.826751255617235e-05,
+      "loss": 1.4131,
+      "step": 6564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2323434352874756,
+      "learning_rate": 4.826724821570183e-05,
+      "loss": 1.6668,
+      "step": 6565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0093281269073486,
+      "learning_rate": 4.8266983875231295e-05,
+      "loss": 1.4954,
+      "step": 6566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4359073638916016,
+      "learning_rate": 4.8266719534760775e-05,
+      "loss": 1.5358,
+      "step": 6567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8539221286773682,
+      "learning_rate": 4.826645519429025e-05,
+      "loss": 1.3895,
+      "step": 6568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.56770658493042,
+      "learning_rate": 4.826619085381972e-05,
+      "loss": 1.312,
+      "step": 6569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.766993522644043,
+      "learning_rate": 4.8265926513349196e-05,
+      "loss": 1.2658,
+      "step": 6570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.180316686630249,
+      "learning_rate": 4.826566217287867e-05,
+      "loss": 1.5599,
+      "step": 6571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7363998889923096,
+      "learning_rate": 4.826539783240814e-05,
+      "loss": 1.5863,
+      "step": 6572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3143893480300903,
+      "learning_rate": 4.8265133491937617e-05,
+      "loss": 1.4395,
+      "step": 6573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0776278972625732,
+      "learning_rate": 4.826486915146709e-05,
+      "loss": 1.4983,
+      "step": 6574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7437769174575806,
+      "learning_rate": 4.8264604810996564e-05,
+      "loss": 1.481,
+      "step": 6575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7049026489257812,
+      "learning_rate": 4.826434047052604e-05,
+      "loss": 1.4684,
+      "step": 6576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7246581315994263,
+      "learning_rate": 4.826407613005551e-05,
+      "loss": 1.2792,
+      "step": 6577
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8020905256271362,
+      "learning_rate": 4.826381178958499e-05,
+      "loss": 1.4944,
+      "step": 6578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3356244564056396,
+      "learning_rate": 4.826354744911446e-05,
+      "loss": 1.8695,
+      "step": 6579
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4350972175598145,
+      "learning_rate": 4.826328310864394e-05,
+      "loss": 1.5366,
+      "step": 6580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8342818021774292,
+      "learning_rate": 4.8263018768173405e-05,
+      "loss": 1.357,
+      "step": 6581
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8791284561157227,
+      "learning_rate": 4.8262754427702885e-05,
+      "loss": 1.4181,
+      "step": 6582
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4840872287750244,
+      "learning_rate": 4.826249008723236e-05,
+      "loss": 1.563,
+      "step": 6583
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.571995496749878,
+      "learning_rate": 4.826222574676183e-05,
+      "loss": 1.4945,
+      "step": 6584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.656173825263977,
+      "learning_rate": 4.8261961406291306e-05,
+      "loss": 1.414,
+      "step": 6585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7580574750900269,
+      "learning_rate": 4.826169706582078e-05,
+      "loss": 1.5249,
+      "step": 6586
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6603020429611206,
+      "learning_rate": 4.826143272535025e-05,
+      "loss": 1.4615,
+      "step": 6587
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8886467218399048,
+      "learning_rate": 4.826116838487973e-05,
+      "loss": 1.3731,
+      "step": 6588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3595576286315918,
+      "learning_rate": 4.82609040444092e-05,
+      "loss": 1.2579,
+      "step": 6589
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5509822368621826,
+      "learning_rate": 4.8260639703938674e-05,
+      "loss": 1.3204,
+      "step": 6590
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.955670118331909,
+      "learning_rate": 4.826037536346815e-05,
+      "loss": 1.5244,
+      "step": 6591
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.568281888961792,
+      "learning_rate": 4.826011102299762e-05,
+      "loss": 1.3742,
+      "step": 6592
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3107614517211914,
+      "learning_rate": 4.82598466825271e-05,
+      "loss": 1.4056,
+      "step": 6593
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6090508699417114,
+      "learning_rate": 4.825958234205657e-05,
+      "loss": 1.3918,
+      "step": 6594
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.8016252517700195,
+      "learning_rate": 4.825931800158605e-05,
+      "loss": 1.2728,
+      "step": 6595
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7302873134613037,
+      "learning_rate": 4.8259053661115515e-05,
+      "loss": 1.4615,
+      "step": 6596
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2375612258911133,
+      "learning_rate": 4.8258789320644996e-05,
+      "loss": 1.4504,
+      "step": 6597
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7467232942581177,
+      "learning_rate": 4.825852498017447e-05,
+      "loss": 2.073,
+      "step": 6598
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5889360904693604,
+      "learning_rate": 4.8258260639703936e-05,
+      "loss": 2.3324,
+      "step": 6599
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4541959762573242,
+      "learning_rate": 4.8257996299233417e-05,
+      "loss": 1.3601,
+      "step": 6600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5147041082382202,
+      "learning_rate": 4.825773195876288e-05,
+      "loss": 1.3451,
+      "step": 6601
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.120753288269043,
+      "learning_rate": 4.8257467618292364e-05,
+      "loss": 1.5383,
+      "step": 6602
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7569901943206787,
+      "learning_rate": 4.825720327782184e-05,
+      "loss": 1.5654,
+      "step": 6603
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7337852716445923,
+      "learning_rate": 4.825693893735131e-05,
+      "loss": 1.4208,
+      "step": 6604
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.65287446975708,
+      "learning_rate": 4.8256674596880784e-05,
+      "loss": 2.1072,
+      "step": 6605
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.707714319229126,
+      "learning_rate": 4.825641025641026e-05,
+      "loss": 1.4584,
+      "step": 6606
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9081634283065796,
+      "learning_rate": 4.825614591593973e-05,
+      "loss": 1.4823,
+      "step": 6607
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6446962356567383,
+      "learning_rate": 4.8255881575469205e-05,
+      "loss": 1.3507,
+      "step": 6608
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.838681936264038,
+      "learning_rate": 4.825561723499868e-05,
+      "loss": 1.4744,
+      "step": 6609
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.051640033721924,
+      "learning_rate": 4.825535289452815e-05,
+      "loss": 1.4026,
+      "step": 6610
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.774850845336914,
+      "learning_rate": 4.8255088554057626e-05,
+      "loss": 1.5412,
+      "step": 6611
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6094833612442017,
+      "learning_rate": 4.82548242135871e-05,
+      "loss": 1.397,
+      "step": 6612
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4261589050292969,
+      "learning_rate": 4.825455987311658e-05,
+      "loss": 1.264,
+      "step": 6613
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7102460861206055,
+      "learning_rate": 4.8254295532646047e-05,
+      "loss": 1.4124,
+      "step": 6614
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6939876079559326,
+      "learning_rate": 4.825403119217553e-05,
+      "loss": 1.3958,
+      "step": 6615
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8743345737457275,
+      "learning_rate": 4.8253766851704994e-05,
+      "loss": 1.5768,
+      "step": 6616
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5334484577178955,
+      "learning_rate": 4.8253502511234474e-05,
+      "loss": 1.8414,
+      "step": 6617
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7456269264221191,
+      "learning_rate": 4.825323817076395e-05,
+      "loss": 1.3279,
+      "step": 6618
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.064530849456787,
+      "learning_rate": 4.825297383029342e-05,
+      "loss": 1.3578,
+      "step": 6619
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5948501825332642,
+      "learning_rate": 4.8252709489822895e-05,
+      "loss": 1.1947,
+      "step": 6620
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8802729845046997,
+      "learning_rate": 4.825244514935237e-05,
+      "loss": 1.4181,
+      "step": 6621
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8933850526809692,
+      "learning_rate": 4.825218080888184e-05,
+      "loss": 1.4568,
+      "step": 6622
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1007847785949707,
+      "learning_rate": 4.8251916468411315e-05,
+      "loss": 1.3959,
+      "step": 6623
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7617509365081787,
+      "learning_rate": 4.825165212794079e-05,
+      "loss": 1.418,
+      "step": 6624
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7610834836959839,
+      "learning_rate": 4.825138778747026e-05,
+      "loss": 1.4008,
+      "step": 6625
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8488237857818604,
+      "learning_rate": 4.8251123446999736e-05,
+      "loss": 1.6118,
+      "step": 6626
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7420222759246826,
+      "learning_rate": 4.825085910652921e-05,
+      "loss": 1.3133,
+      "step": 6627
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.160501003265381,
+      "learning_rate": 4.825059476605869e-05,
+      "loss": 1.5051,
+      "step": 6628
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6764954328536987,
+      "learning_rate": 4.825033042558816e-05,
+      "loss": 1.3555,
+      "step": 6629
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3801004886627197,
+      "learning_rate": 4.825006608511764e-05,
+      "loss": 1.1218,
+      "step": 6630
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.899067759513855,
+      "learning_rate": 4.8249801744647104e-05,
+      "loss": 1.3324,
+      "step": 6631
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6317296028137207,
+      "learning_rate": 4.8249537404176584e-05,
+      "loss": 1.3942,
+      "step": 6632
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4298748970031738,
+      "learning_rate": 4.824927306370606e-05,
+      "loss": 1.314,
+      "step": 6633
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9906808137893677,
+      "learning_rate": 4.824900872323553e-05,
+      "loss": 1.5296,
+      "step": 6634
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5967448949813843,
+      "learning_rate": 4.8248744382765005e-05,
+      "loss": 1.2252,
+      "step": 6635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7575392723083496,
+      "learning_rate": 4.824848004229447e-05,
+      "loss": 1.3569,
+      "step": 6636
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0080924034118652,
+      "learning_rate": 4.824821570182395e-05,
+      "loss": 2.0964,
+      "step": 6637
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0378146171569824,
+      "learning_rate": 4.8247951361353426e-05,
+      "loss": 1.4938,
+      "step": 6638
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6722849607467651,
+      "learning_rate": 4.82476870208829e-05,
+      "loss": 1.4174,
+      "step": 6639
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0503780841827393,
+      "learning_rate": 4.824742268041237e-05,
+      "loss": 1.4972,
+      "step": 6640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4831526279449463,
+      "learning_rate": 4.8247158339941846e-05,
+      "loss": 1.2754,
+      "step": 6641
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5943167209625244,
+      "learning_rate": 4.824689399947132e-05,
+      "loss": 1.3486,
+      "step": 6642
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5922707319259644,
+      "learning_rate": 4.82466296590008e-05,
+      "loss": 1.3462,
+      "step": 6643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7302312850952148,
+      "learning_rate": 4.824636531853027e-05,
+      "loss": 1.333,
+      "step": 6644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3600311279296875,
+      "learning_rate": 4.824610097805974e-05,
+      "loss": 1.2698,
+      "step": 6645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.004925012588501,
+      "learning_rate": 4.8245836637589214e-05,
+      "loss": 1.4788,
+      "step": 6646
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8024418354034424,
+      "learning_rate": 4.824557229711869e-05,
+      "loss": 1.5199,
+      "step": 6647
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8858249187469482,
+      "learning_rate": 4.824530795664817e-05,
+      "loss": 1.5249,
+      "step": 6648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1211135387420654,
+      "learning_rate": 4.8245043616177635e-05,
+      "loss": 1.4221,
+      "step": 6649
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6887212991714478,
+      "learning_rate": 4.8244779275707115e-05,
+      "loss": 1.3322,
+      "step": 6650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9310022592544556,
+      "learning_rate": 4.824451493523658e-05,
+      "loss": 1.4128,
+      "step": 6651
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.603998064994812,
+      "learning_rate": 4.824425059476606e-05,
+      "loss": 1.3638,
+      "step": 6652
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2772178649902344,
+      "learning_rate": 4.8243986254295536e-05,
+      "loss": 1.3712,
+      "step": 6653
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.794693112373352,
+      "learning_rate": 4.824372191382501e-05,
+      "loss": 1.4334,
+      "step": 6654
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8565353155136108,
+      "learning_rate": 4.824345757335448e-05,
+      "loss": 1.3107,
+      "step": 6655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5430495738983154,
+      "learning_rate": 4.824319323288396e-05,
+      "loss": 1.4092,
+      "step": 6656
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.220773458480835,
+      "learning_rate": 4.824292889241343e-05,
+      "loss": 1.5302,
+      "step": 6657
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5505123138427734,
+      "learning_rate": 4.8242664551942904e-05,
+      "loss": 1.3059,
+      "step": 6658
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7606844902038574,
+      "learning_rate": 4.824240021147238e-05,
+      "loss": 1.3589,
+      "step": 6659
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5247254371643066,
+      "learning_rate": 4.824213587100185e-05,
+      "loss": 1.2895,
+      "step": 6660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3901171684265137,
+      "learning_rate": 4.8241871530531325e-05,
+      "loss": 1.4191,
+      "step": 6661
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8800420761108398,
+      "learning_rate": 4.82416071900608e-05,
+      "loss": 1.4259,
+      "step": 6662
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.859657049179077,
+      "learning_rate": 4.824134284959028e-05,
+      "loss": 1.5368,
+      "step": 6663
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.810564637184143,
+      "learning_rate": 4.8241078509119745e-05,
+      "loss": 1.3207,
+      "step": 6664
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8096864223480225,
+      "learning_rate": 4.8240814168649226e-05,
+      "loss": 1.4402,
+      "step": 6665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2123348712921143,
+      "learning_rate": 4.824054982817869e-05,
+      "loss": 1.4976,
+      "step": 6666
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5114413499832153,
+      "learning_rate": 4.824028548770817e-05,
+      "loss": 1.5131,
+      "step": 6667
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4915778636932373,
+      "learning_rate": 4.8240021147237646e-05,
+      "loss": 1.3604,
+      "step": 6668
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2583965063095093,
+      "learning_rate": 4.823975680676712e-05,
+      "loss": 1.2327,
+      "step": 6669
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.33100163936615,
+      "learning_rate": 4.8239492466296594e-05,
+      "loss": 1.3036,
+      "step": 6670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.600347638130188,
+      "learning_rate": 4.823922812582606e-05,
+      "loss": 1.4006,
+      "step": 6671
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.196754217147827,
+      "learning_rate": 4.823896378535554e-05,
+      "loss": 1.472,
+      "step": 6672
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6722278594970703,
+      "learning_rate": 4.8238699444885014e-05,
+      "loss": 1.5599,
+      "step": 6673
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.724501371383667,
+      "learning_rate": 4.823843510441449e-05,
+      "loss": 1.4136,
+      "step": 6674
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6867225170135498,
+      "learning_rate": 4.823817076394396e-05,
+      "loss": 1.3556,
+      "step": 6675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8065874576568604,
+      "learning_rate": 4.8237906423473435e-05,
+      "loss": 1.4917,
+      "step": 6676
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.194286346435547,
+      "learning_rate": 4.823764208300291e-05,
+      "loss": 1.6821,
+      "step": 6677
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8574241399765015,
+      "learning_rate": 4.823737774253239e-05,
+      "loss": 1.5266,
+      "step": 6678
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.001945972442627,
+      "learning_rate": 4.8237113402061856e-05,
+      "loss": 1.3285,
+      "step": 6679
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6860787868499756,
+      "learning_rate": 4.823684906159133e-05,
+      "loss": 1.4015,
+      "step": 6680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.548853874206543,
+      "learning_rate": 4.82365847211208e-05,
+      "loss": 1.3001,
+      "step": 6681
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.676652431488037,
+      "learning_rate": 4.8236320380650276e-05,
+      "loss": 1.4904,
+      "step": 6682
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.482914924621582,
+      "learning_rate": 4.823605604017976e-05,
+      "loss": 1.3065,
+      "step": 6683
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8636629581451416,
+      "learning_rate": 4.8235791699709224e-05,
+      "loss": 1.3923,
+      "step": 6684
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.711036443710327,
+      "learning_rate": 4.8235527359238704e-05,
+      "loss": 1.5534,
+      "step": 6685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.591152548789978,
+      "learning_rate": 4.823526301876817e-05,
+      "loss": 1.2176,
+      "step": 6686
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1628403663635254,
+      "learning_rate": 4.823499867829765e-05,
+      "loss": 1.4135,
+      "step": 6687
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.788426399230957,
+      "learning_rate": 4.8234734337827125e-05,
+      "loss": 1.482,
+      "step": 6688
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7718288898468018,
+      "learning_rate": 4.82344699973566e-05,
+      "loss": 1.251,
+      "step": 6689
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6099847555160522,
+      "learning_rate": 4.823420565688607e-05,
+      "loss": 1.753,
+      "step": 6690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.85465407371521,
+      "learning_rate": 4.8233941316415545e-05,
+      "loss": 1.4386,
+      "step": 6691
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8299190998077393,
+      "learning_rate": 4.823367697594502e-05,
+      "loss": 1.3453,
+      "step": 6692
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8316762447357178,
+      "learning_rate": 4.823341263547449e-05,
+      "loss": 1.4316,
+      "step": 6693
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.826007604598999,
+      "learning_rate": 4.8233148295003966e-05,
+      "loss": 1.3108,
+      "step": 6694
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.781666874885559,
+      "learning_rate": 4.823288395453344e-05,
+      "loss": 1.4727,
+      "step": 6695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.152873992919922,
+      "learning_rate": 4.823261961406291e-05,
+      "loss": 1.4016,
+      "step": 6696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0739822387695312,
+      "learning_rate": 4.823235527359239e-05,
+      "loss": 1.2868,
+      "step": 6697
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.210247278213501,
+      "learning_rate": 4.823209093312187e-05,
+      "loss": 1.3434,
+      "step": 6698
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9673181772232056,
+      "learning_rate": 4.8231826592651334e-05,
+      "loss": 1.5078,
+      "step": 6699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.387813091278076,
+      "learning_rate": 4.8231562252180814e-05,
+      "loss": 1.4816,
+      "step": 6700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.168231248855591,
+      "learning_rate": 4.823129791171028e-05,
+      "loss": 1.2645,
+      "step": 6701
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6742557287216187,
+      "learning_rate": 4.823103357123976e-05,
+      "loss": 1.3003,
+      "step": 6702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.882627010345459,
+      "learning_rate": 4.8230769230769235e-05,
+      "loss": 1.4948,
+      "step": 6703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4613351821899414,
+      "learning_rate": 4.823050489029871e-05,
+      "loss": 1.3996,
+      "step": 6704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6211466789245605,
+      "learning_rate": 4.823024054982818e-05,
+      "loss": 1.3139,
+      "step": 6705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9584745168685913,
+      "learning_rate": 4.8229976209357656e-05,
+      "loss": 1.5155,
+      "step": 6706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.373211145401001,
+      "learning_rate": 4.822971186888713e-05,
+      "loss": 1.5675,
+      "step": 6707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5229337215423584,
+      "learning_rate": 4.82294475284166e-05,
+      "loss": 1.5782,
+      "step": 6708
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.228278636932373,
+      "learning_rate": 4.8229183187946076e-05,
+      "loss": 1.4974,
+      "step": 6709
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8478854894638062,
+      "learning_rate": 4.822891884747555e-05,
+      "loss": 1.4184,
+      "step": 6710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.914621353149414,
+      "learning_rate": 4.8228654507005024e-05,
+      "loss": 1.5302,
+      "step": 6711
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5413990020751953,
+      "learning_rate": 4.82283901665345e-05,
+      "loss": 1.3809,
+      "step": 6712
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6454380750656128,
+      "learning_rate": 4.822812582606398e-05,
+      "loss": 1.4893,
+      "step": 6713
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.724995493888855,
+      "learning_rate": 4.8227861485593444e-05,
+      "loss": 1.3705,
+      "step": 6714
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9537824392318726,
+      "learning_rate": 4.8227597145122925e-05,
+      "loss": 1.5092,
+      "step": 6715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7390862703323364,
+      "learning_rate": 4.822733280465239e-05,
+      "loss": 1.3434,
+      "step": 6716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7581617832183838,
+      "learning_rate": 4.8227068464181865e-05,
+      "loss": 1.3716,
+      "step": 6717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9375813007354736,
+      "learning_rate": 4.8226804123711345e-05,
+      "loss": 1.3807,
+      "step": 6718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5088646411895752,
+      "learning_rate": 4.822653978324081e-05,
+      "loss": 1.326,
+      "step": 6719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7618147134780884,
+      "learning_rate": 4.822627544277029e-05,
+      "loss": 1.4775,
+      "step": 6720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.954552412033081,
+      "learning_rate": 4.822601110229976e-05,
+      "loss": 1.4563,
+      "step": 6721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.549358367919922,
+      "learning_rate": 4.822574676182924e-05,
+      "loss": 1.46,
+      "step": 6722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6343919038772583,
+      "learning_rate": 4.822548242135871e-05,
+      "loss": 1.3727,
+      "step": 6723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9067964553833008,
+      "learning_rate": 4.822521808088819e-05,
+      "loss": 1.4992,
+      "step": 6724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9555072784423828,
+      "learning_rate": 4.822495374041766e-05,
+      "loss": 1.5236,
+      "step": 6725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4759035110473633,
+      "learning_rate": 4.8224689399947134e-05,
+      "loss": 1.365,
+      "step": 6726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9179940223693848,
+      "learning_rate": 4.822442505947661e-05,
+      "loss": 1.5255,
+      "step": 6727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8269222974777222,
+      "learning_rate": 4.822416071900608e-05,
+      "loss": 1.418,
+      "step": 6728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1328978538513184,
+      "learning_rate": 4.8223896378535555e-05,
+      "loss": 1.4733,
+      "step": 6729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7149276733398438,
+      "learning_rate": 4.822363203806503e-05,
+      "loss": 1.4096,
+      "step": 6730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.898821473121643,
+      "learning_rate": 4.82233676975945e-05,
+      "loss": 1.5619,
+      "step": 6731
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8014065027236938,
+      "learning_rate": 4.8223103357123975e-05,
+      "loss": 1.3323,
+      "step": 6732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9247713088989258,
+      "learning_rate": 4.8222839016653456e-05,
+      "loss": 1.4162,
+      "step": 6733
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3599423170089722,
+      "learning_rate": 4.822257467618292e-05,
+      "loss": 1.284,
+      "step": 6734
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8829020261764526,
+      "learning_rate": 4.82223103357124e-05,
+      "loss": 1.3694,
+      "step": 6735
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.133199691772461,
+      "learning_rate": 4.822204599524187e-05,
+      "loss": 1.4459,
+      "step": 6736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8177059888839722,
+      "learning_rate": 4.822178165477135e-05,
+      "loss": 1.6399,
+      "step": 6737
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7157459259033203,
+      "learning_rate": 4.8221517314300824e-05,
+      "loss": 1.3699,
+      "step": 6738
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7175670862197876,
+      "learning_rate": 4.82212529738303e-05,
+      "loss": 1.3465,
+      "step": 6739
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9629031419754028,
+      "learning_rate": 4.822098863335977e-05,
+      "loss": 1.3888,
+      "step": 6740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.788087010383606,
+      "learning_rate": 4.8220724292889244e-05,
+      "loss": 1.3647,
+      "step": 6741
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6764038801193237,
+      "learning_rate": 4.822045995241872e-05,
+      "loss": 1.1467,
+      "step": 6742
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4352415800094604,
+      "learning_rate": 4.822019561194819e-05,
+      "loss": 1.4562,
+      "step": 6743
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6587698459625244,
+      "learning_rate": 4.8219931271477665e-05,
+      "loss": 1.3845,
+      "step": 6744
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6883310079574585,
+      "learning_rate": 4.821966693100714e-05,
+      "loss": 1.3258,
+      "step": 6745
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8767255544662476,
+      "learning_rate": 4.821940259053661e-05,
+      "loss": 2.1025,
+      "step": 6746
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9841525554656982,
+      "learning_rate": 4.8219138250066086e-05,
+      "loss": 1.3321,
+      "step": 6747
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.084047317504883,
+      "learning_rate": 4.8218873909595566e-05,
+      "loss": 1.4606,
+      "step": 6748
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.418532848358154,
+      "learning_rate": 4.821860956912503e-05,
+      "loss": 1.6478,
+      "step": 6749
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5554016828536987,
+      "learning_rate": 4.821834522865451e-05,
+      "loss": 1.3473,
+      "step": 6750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.379040002822876,
+      "learning_rate": 4.821808088818398e-05,
+      "loss": 1.6736,
+      "step": 6751
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9192800521850586,
+      "learning_rate": 4.821781654771346e-05,
+      "loss": 1.3026,
+      "step": 6752
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5902564525604248,
+      "learning_rate": 4.8217552207242934e-05,
+      "loss": 1.3657,
+      "step": 6753
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.635662317276001,
+      "learning_rate": 4.82172878667724e-05,
+      "loss": 1.4104,
+      "step": 6754
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.406409502029419,
+      "learning_rate": 4.821702352630188e-05,
+      "loss": 1.349,
+      "step": 6755
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1364104747772217,
+      "learning_rate": 4.821675918583135e-05,
+      "loss": 1.4763,
+      "step": 6756
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8173112869262695,
+      "learning_rate": 4.821649484536083e-05,
+      "loss": 1.5561,
+      "step": 6757
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6839903593063354,
+      "learning_rate": 4.82162305048903e-05,
+      "loss": 1.5047,
+      "step": 6758
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8539842367172241,
+      "learning_rate": 4.8215966164419775e-05,
+      "loss": 1.5084,
+      "step": 6759
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8488625288009644,
+      "learning_rate": 4.821570182394925e-05,
+      "loss": 1.4272,
+      "step": 6760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6597309112548828,
+      "learning_rate": 4.821543748347872e-05,
+      "loss": 1.4561,
+      "step": 6761
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5841479301452637,
+      "learning_rate": 4.8215173143008196e-05,
+      "loss": 1.3873,
+      "step": 6762
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9843717813491821,
+      "learning_rate": 4.821490880253767e-05,
+      "loss": 1.2643,
+      "step": 6763
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7158598899841309,
+      "learning_rate": 4.821464446206714e-05,
+      "loss": 1.5147,
+      "step": 6764
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.129439115524292,
+      "learning_rate": 4.821438012159662e-05,
+      "loss": 1.5193,
+      "step": 6765
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7988207340240479,
+      "learning_rate": 4.821411578112609e-05,
+      "loss": 1.4976,
+      "step": 6766
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9958490133285522,
+      "learning_rate": 4.8213851440655564e-05,
+      "loss": 1.5924,
+      "step": 6767
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.73973548412323,
+      "learning_rate": 4.8213587100185044e-05,
+      "loss": 1.4418,
+      "step": 6768
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2767462730407715,
+      "learning_rate": 4.821332275971451e-05,
+      "loss": 1.5265,
+      "step": 6769
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4522449970245361,
+      "learning_rate": 4.821305841924399e-05,
+      "loss": 1.2747,
+      "step": 6770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5302042961120605,
+      "learning_rate": 4.821279407877346e-05,
+      "loss": 1.2692,
+      "step": 6771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.754840612411499,
+      "learning_rate": 4.821252973830294e-05,
+      "loss": 1.5051,
+      "step": 6772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8306812047958374,
+      "learning_rate": 4.821226539783241e-05,
+      "loss": 1.5157,
+      "step": 6773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.978913426399231,
+      "learning_rate": 4.8212001057361886e-05,
+      "loss": 1.5173,
+      "step": 6774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7886606454849243,
+      "learning_rate": 4.821173671689136e-05,
+      "loss": 1.3466,
+      "step": 6775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.304089069366455,
+      "learning_rate": 4.821147237642083e-05,
+      "loss": 1.5479,
+      "step": 6776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.676723599433899,
+      "learning_rate": 4.8211208035950306e-05,
+      "loss": 1.4515,
+      "step": 6777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6664212942123413,
+      "learning_rate": 4.821094369547978e-05,
+      "loss": 1.4107,
+      "step": 6778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6482583284378052,
+      "learning_rate": 4.8210679355009253e-05,
+      "loss": 1.3642,
+      "step": 6779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0200629234313965,
+      "learning_rate": 4.821041501453873e-05,
+      "loss": 1.5814,
+      "step": 6780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.206848382949829,
+      "learning_rate": 4.82101506740682e-05,
+      "loss": 1.5043,
+      "step": 6781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6113061904907227,
+      "learning_rate": 4.8209886333597674e-05,
+      "loss": 1.4619,
+      "step": 6782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8279345035552979,
+      "learning_rate": 4.8209621993127155e-05,
+      "loss": 1.3788,
+      "step": 6783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7288391590118408,
+      "learning_rate": 4.820935765265662e-05,
+      "loss": 1.4032,
+      "step": 6784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8135991096496582,
+      "learning_rate": 4.82090933121861e-05,
+      "loss": 1.3579,
+      "step": 6785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5332999229431152,
+      "learning_rate": 4.820882897171557e-05,
+      "loss": 1.4019,
+      "step": 6786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5625176429748535,
+      "learning_rate": 4.820856463124505e-05,
+      "loss": 1.332,
+      "step": 6787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5822092294692993,
+      "learning_rate": 4.820830029077452e-05,
+      "loss": 1.433,
+      "step": 6788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.701380968093872,
+      "learning_rate": 4.820803595030399e-05,
+      "loss": 1.3219,
+      "step": 6789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.185961961746216,
+      "learning_rate": 4.820777160983347e-05,
+      "loss": 1.5548,
+      "step": 6790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5697836875915527,
+      "learning_rate": 4.8207507269362936e-05,
+      "loss": 1.3321,
+      "step": 6791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5730094909667969,
+      "learning_rate": 4.820724292889242e-05,
+      "loss": 1.3133,
+      "step": 6792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5201021432876587,
+      "learning_rate": 4.820697858842189e-05,
+      "loss": 1.4438,
+      "step": 6793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3393160104751587,
+      "learning_rate": 4.8206714247951364e-05,
+      "loss": 1.2297,
+      "step": 6794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3271632194519043,
+      "learning_rate": 4.820644990748084e-05,
+      "loss": 1.4981,
+      "step": 6795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4270941019058228,
+      "learning_rate": 4.820618556701031e-05,
+      "loss": 1.191,
+      "step": 6796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.622985601425171,
+      "learning_rate": 4.8205921226539785e-05,
+      "loss": 1.2663,
+      "step": 6797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9148688316345215,
+      "learning_rate": 4.820565688606926e-05,
+      "loss": 1.4844,
+      "step": 6798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.196896553039551,
+      "learning_rate": 4.820539254559873e-05,
+      "loss": 1.3495,
+      "step": 6799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7246558666229248,
+      "learning_rate": 4.8205128205128205e-05,
+      "loss": 1.383,
+      "step": 6800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.224186658859253,
+      "learning_rate": 4.820486386465768e-05,
+      "loss": 1.5184,
+      "step": 6801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.218459129333496,
+      "learning_rate": 4.820459952418715e-05,
+      "loss": 1.5373,
+      "step": 6802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3243407011032104,
+      "learning_rate": 4.820433518371663e-05,
+      "loss": 1.3313,
+      "step": 6803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.813349723815918,
+      "learning_rate": 4.82040708432461e-05,
+      "loss": 2.4161,
+      "step": 6804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7155495882034302,
+      "learning_rate": 4.820380650277558e-05,
+      "loss": 1.309,
+      "step": 6805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8316320180892944,
+      "learning_rate": 4.820354216230505e-05,
+      "loss": 1.3867,
+      "step": 6806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.980057716369629,
+      "learning_rate": 4.820327782183453e-05,
+      "loss": 1.465,
+      "step": 6807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9099912643432617,
+      "learning_rate": 4.8203013481364e-05,
+      "loss": 1.456,
+      "step": 6808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8151378631591797,
+      "learning_rate": 4.8202749140893474e-05,
+      "loss": 1.3353,
+      "step": 6809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.864927053451538,
+      "learning_rate": 4.820248480042295e-05,
+      "loss": 1.2725,
+      "step": 6810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.425016164779663,
+      "learning_rate": 4.820222045995242e-05,
+      "loss": 1.5486,
+      "step": 6811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7455166578292847,
+      "learning_rate": 4.8201956119481895e-05,
+      "loss": 1.4913,
+      "step": 6812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8460808992385864,
+      "learning_rate": 4.820169177901137e-05,
+      "loss": 1.4564,
+      "step": 6813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5313783884048462,
+      "learning_rate": 4.820142743854084e-05,
+      "loss": 1.4169,
+      "step": 6814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.551130533218384,
+      "learning_rate": 4.8201163098070316e-05,
+      "loss": 1.5513,
+      "step": 6815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3388590812683105,
+      "learning_rate": 4.820089875759979e-05,
+      "loss": 1.5594,
+      "step": 6816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8769012689590454,
+      "learning_rate": 4.820063441712926e-05,
+      "loss": 1.3187,
+      "step": 6817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7618584632873535,
+      "learning_rate": 4.820037007665874e-05,
+      "loss": 1.2611,
+      "step": 6818
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7105064392089844,
+      "learning_rate": 4.820010573618821e-05,
+      "loss": 1.5337,
+      "step": 6819
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.214426279067993,
+      "learning_rate": 4.819984139571769e-05,
+      "loss": 1.3,
+      "step": 6820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7881313562393188,
+      "learning_rate": 4.819957705524716e-05,
+      "loss": 1.4785,
+      "step": 6821
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4660691022872925,
+      "learning_rate": 4.819931271477664e-05,
+      "loss": 1.1834,
+      "step": 6822
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5670647621154785,
+      "learning_rate": 4.819904837430611e-05,
+      "loss": 1.2426,
+      "step": 6823
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3523502349853516,
+      "learning_rate": 4.8198784033835585e-05,
+      "loss": 1.3843,
+      "step": 6824
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.70943021774292,
+      "learning_rate": 4.819851969336506e-05,
+      "loss": 1.5074,
+      "step": 6825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7319227457046509,
+      "learning_rate": 4.8198255352894525e-05,
+      "loss": 1.476,
+      "step": 6826
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0385122299194336,
+      "learning_rate": 4.8197991012424005e-05,
+      "loss": 1.5053,
+      "step": 6827
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7466739416122437,
+      "learning_rate": 4.819772667195348e-05,
+      "loss": 1.4071,
+      "step": 6828
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7974426746368408,
+      "learning_rate": 4.819746233148295e-05,
+      "loss": 1.1857,
+      "step": 6829
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7845615148544312,
+      "learning_rate": 4.8197197991012426e-05,
+      "loss": 1.4348,
+      "step": 6830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8993710279464722,
+      "learning_rate": 4.81969336505419e-05,
+      "loss": 1.3924,
+      "step": 6831
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.609743356704712,
+      "learning_rate": 4.819666931007137e-05,
+      "loss": 1.4666,
+      "step": 6832
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8062622547149658,
+      "learning_rate": 4.8196404969600853e-05,
+      "loss": 1.2989,
+      "step": 6833
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3045685291290283,
+      "learning_rate": 4.819614062913032e-05,
+      "loss": 1.549,
+      "step": 6834
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6319725513458252,
+      "learning_rate": 4.8195876288659794e-05,
+      "loss": 2.4619,
+      "step": 6835
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0392792224884033,
+      "learning_rate": 4.819561194818927e-05,
+      "loss": 2.2238,
+      "step": 6836
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7144312858581543,
+      "learning_rate": 4.819534760771874e-05,
+      "loss": 1.4099,
+      "step": 6837
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9040653705596924,
+      "learning_rate": 4.819508326724822e-05,
+      "loss": 1.5216,
+      "step": 6838
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8126404285430908,
+      "learning_rate": 4.819481892677769e-05,
+      "loss": 1.3051,
+      "step": 6839
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.303869605064392,
+      "learning_rate": 4.819455458630717e-05,
+      "loss": 1.2287,
+      "step": 6840
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4013986587524414,
+      "learning_rate": 4.8194290245836635e-05,
+      "loss": 1.5353,
+      "step": 6841
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.412905216217041,
+      "learning_rate": 4.8194025905366116e-05,
+      "loss": 1.596,
+      "step": 6842
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0264525413513184,
+      "learning_rate": 4.819376156489559e-05,
+      "loss": 1.7466,
+      "step": 6843
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.599138855934143,
+      "learning_rate": 4.819349722442506e-05,
+      "loss": 1.3859,
+      "step": 6844
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.749320387840271,
+      "learning_rate": 4.8193232883954536e-05,
+      "loss": 1.3615,
+      "step": 6845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7781983613967896,
+      "learning_rate": 4.819296854348401e-05,
+      "loss": 1.4864,
+      "step": 6846
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1943252086639404,
+      "learning_rate": 4.8192704203013483e-05,
+      "loss": 1.5686,
+      "step": 6847
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.590442419052124,
+      "learning_rate": 4.819243986254296e-05,
+      "loss": 1.324,
+      "step": 6848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.000605344772339,
+      "learning_rate": 4.819217552207243e-05,
+      "loss": 1.4631,
+      "step": 6849
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.918105602264404,
+      "learning_rate": 4.8191911181601904e-05,
+      "loss": 2.0344,
+      "step": 6850
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.73543119430542,
+      "learning_rate": 4.819164684113138e-05,
+      "loss": 1.4264,
+      "step": 6851
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1430301666259766,
+      "learning_rate": 4.819138250066085e-05,
+      "loss": 1.4277,
+      "step": 6852
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.314436435699463,
+      "learning_rate": 4.819111816019033e-05,
+      "loss": 1.3745,
+      "step": 6853
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7817548513412476,
+      "learning_rate": 4.81908538197198e-05,
+      "loss": 1.4505,
+      "step": 6854
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6361972093582153,
+      "learning_rate": 4.819058947924928e-05,
+      "loss": 1.7377,
+      "step": 6855
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.590302586555481,
+      "learning_rate": 4.8190325138778746e-05,
+      "loss": 1.4945,
+      "step": 6856
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7460488080978394,
+      "learning_rate": 4.8190060798308226e-05,
+      "loss": 1.3176,
+      "step": 6857
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.904443383216858,
+      "learning_rate": 4.81897964578377e-05,
+      "loss": 1.5368,
+      "step": 6858
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8913196325302124,
+      "learning_rate": 4.818953211736717e-05,
+      "loss": 1.5125,
+      "step": 6859
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7731674909591675,
+      "learning_rate": 4.8189267776896647e-05,
+      "loss": 1.462,
+      "step": 6860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.032395601272583,
+      "learning_rate": 4.8189003436426113e-05,
+      "loss": 1.4485,
+      "step": 6861
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9787179231643677,
+      "learning_rate": 4.8188739095955594e-05,
+      "loss": 1.7676,
+      "step": 6862
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4096564054489136,
+      "learning_rate": 4.818847475548507e-05,
+      "loss": 1.4138,
+      "step": 6863
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9129940271377563,
+      "learning_rate": 4.818821041501454e-05,
+      "loss": 1.4516,
+      "step": 6864
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.168365001678467,
+      "learning_rate": 4.8187946074544014e-05,
+      "loss": 1.4132,
+      "step": 6865
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.870744228363037,
+      "learning_rate": 4.818768173407349e-05,
+      "loss": 1.4841,
+      "step": 6866
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1823885440826416,
+      "learning_rate": 4.818741739360296e-05,
+      "loss": 1.5597,
+      "step": 6867
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1679141521453857,
+      "learning_rate": 4.818715305313244e-05,
+      "loss": 1.3939,
+      "step": 6868
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.071078300476074,
+      "learning_rate": 4.818688871266191e-05,
+      "loss": 1.5885,
+      "step": 6869
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.753304123878479,
+      "learning_rate": 4.818662437219138e-05,
+      "loss": 1.4941,
+      "step": 6870
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5432472229003906,
+      "learning_rate": 4.8186360031720856e-05,
+      "loss": 1.4891,
+      "step": 6871
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.670182466506958,
+      "learning_rate": 4.818609569125033e-05,
+      "loss": 1.4871,
+      "step": 6872
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6292014122009277,
+      "learning_rate": 4.818583135077981e-05,
+      "loss": 1.4779,
+      "step": 6873
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7494271993637085,
+      "learning_rate": 4.8185567010309277e-05,
+      "loss": 1.4595,
+      "step": 6874
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.622512102127075,
+      "learning_rate": 4.818530266983876e-05,
+      "loss": 1.4838,
+      "step": 6875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0830163955688477,
+      "learning_rate": 4.8185038329368224e-05,
+      "loss": 1.331,
+      "step": 6876
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9558629989624023,
+      "learning_rate": 4.8184773988897704e-05,
+      "loss": 1.4299,
+      "step": 6877
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.654399037361145,
+      "learning_rate": 4.818450964842718e-05,
+      "loss": 1.3484,
+      "step": 6878
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9039113521575928,
+      "learning_rate": 4.818424530795665e-05,
+      "loss": 1.4507,
+      "step": 6879
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0992934703826904,
+      "learning_rate": 4.8183980967486125e-05,
+      "loss": 1.4477,
+      "step": 6880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6466342210769653,
+      "learning_rate": 4.81837166270156e-05,
+      "loss": 1.279,
+      "step": 6881
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7685554027557373,
+      "learning_rate": 4.818345228654507e-05,
+      "loss": 1.4769,
+      "step": 6882
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6176362037658691,
+      "learning_rate": 4.8183187946074546e-05,
+      "loss": 1.4752,
+      "step": 6883
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4006391763687134,
+      "learning_rate": 4.818292360560402e-05,
+      "loss": 1.2631,
+      "step": 6884
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8697930574417114,
+      "learning_rate": 4.818265926513349e-05,
+      "loss": 1.4044,
+      "step": 6885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7021851539611816,
+      "learning_rate": 4.8182394924662966e-05,
+      "loss": 1.4145,
+      "step": 6886
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8936532735824585,
+      "learning_rate": 4.818213058419244e-05,
+      "loss": 1.5194,
+      "step": 6887
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9582046270370483,
+      "learning_rate": 4.818186624372192e-05,
+      "loss": 1.3735,
+      "step": 6888
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5924744606018066,
+      "learning_rate": 4.818160190325139e-05,
+      "loss": 1.3558,
+      "step": 6889
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.809525489807129,
+      "learning_rate": 4.818133756278087e-05,
+      "loss": 1.4833,
+      "step": 6890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4661924839019775,
+      "learning_rate": 4.8181073222310334e-05,
+      "loss": 1.5946,
+      "step": 6891
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.012840509414673,
+      "learning_rate": 4.8180808881839814e-05,
+      "loss": 1.5741,
+      "step": 6892
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3162670135498047,
+      "learning_rate": 4.818054454136929e-05,
+      "loss": 1.7315,
+      "step": 6893
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7375813722610474,
+      "learning_rate": 4.818028020089876e-05,
+      "loss": 1.3883,
+      "step": 6894
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6825917959213257,
+      "learning_rate": 4.8180015860428235e-05,
+      "loss": 1.5356,
+      "step": 6895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9199610948562622,
+      "learning_rate": 4.817975151995771e-05,
+      "loss": 1.4819,
+      "step": 6896
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0131587982177734,
+      "learning_rate": 4.817948717948718e-05,
+      "loss": 1.5391,
+      "step": 6897
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6214890480041504,
+      "learning_rate": 4.8179222839016656e-05,
+      "loss": 1.2987,
+      "step": 6898
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6995174884796143,
+      "learning_rate": 4.817895849854613e-05,
+      "loss": 1.3908,
+      "step": 6899
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.66889226436615,
+      "learning_rate": 4.81786941580756e-05,
+      "loss": 1.319,
+      "step": 6900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4697617292404175,
+      "learning_rate": 4.8178429817605077e-05,
+      "loss": 1.2047,
+      "step": 6901
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6001070737838745,
+      "learning_rate": 4.817816547713455e-05,
+      "loss": 1.411,
+      "step": 6902
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7896428108215332,
+      "learning_rate": 4.817790113666403e-05,
+      "loss": 1.47,
+      "step": 6903
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2594714164733887,
+      "learning_rate": 4.81776367961935e-05,
+      "loss": 1.2963,
+      "step": 6904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.303776264190674,
+      "learning_rate": 4.817737245572298e-05,
+      "loss": 1.5995,
+      "step": 6905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6531438827514648,
+      "learning_rate": 4.8177108115252444e-05,
+      "loss": 1.4633,
+      "step": 6906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.953261137008667,
+      "learning_rate": 4.817684377478192e-05,
+      "loss": 1.5585,
+      "step": 6907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8870010375976562,
+      "learning_rate": 4.81765794343114e-05,
+      "loss": 1.4418,
+      "step": 6908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4286311864852905,
+      "learning_rate": 4.8176315093840865e-05,
+      "loss": 1.3368,
+      "step": 6909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7776415348052979,
+      "learning_rate": 4.8176050753370345e-05,
+      "loss": 1.3971,
+      "step": 6910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8171931505203247,
+      "learning_rate": 4.817578641289981e-05,
+      "loss": 1.4308,
+      "step": 6911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1249301433563232,
+      "learning_rate": 4.817552207242929e-05,
+      "loss": 1.4834,
+      "step": 6912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.730483889579773,
+      "learning_rate": 4.8175257731958766e-05,
+      "loss": 1.3512,
+      "step": 6913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6482044458389282,
+      "learning_rate": 4.817499339148824e-05,
+      "loss": 1.5143,
+      "step": 6914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.697542667388916,
+      "learning_rate": 4.817472905101771e-05,
+      "loss": 1.5357,
+      "step": 6915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4032751321792603,
+      "learning_rate": 4.817446471054719e-05,
+      "loss": 2.272,
+      "step": 6916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.336272954940796,
+      "learning_rate": 4.817420037007666e-05,
+      "loss": 1.532,
+      "step": 6917
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3979960680007935,
+      "learning_rate": 4.8173936029606134e-05,
+      "loss": 1.3461,
+      "step": 6918
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.679819941520691,
+      "learning_rate": 4.817367168913561e-05,
+      "loss": 1.3117,
+      "step": 6919
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.497746706008911,
+      "learning_rate": 4.817340734866508e-05,
+      "loss": 1.3942,
+      "step": 6920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7011367082595825,
+      "learning_rate": 4.8173143008194555e-05,
+      "loss": 1.173,
+      "step": 6921
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0999391078948975,
+      "learning_rate": 4.817287866772403e-05,
+      "loss": 1.3109,
+      "step": 6922
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5217264890670776,
+      "learning_rate": 4.817261432725351e-05,
+      "loss": 1.4341,
+      "step": 6923
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4499151706695557,
+      "learning_rate": 4.8172349986782975e-05,
+      "loss": 1.1874,
+      "step": 6924
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7916287183761597,
+      "learning_rate": 4.8172085646312456e-05,
+      "loss": 1.4104,
+      "step": 6925
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6398402452468872,
+      "learning_rate": 4.817182130584192e-05,
+      "loss": 1.4867,
+      "step": 6926
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9428060054779053,
+      "learning_rate": 4.81715569653714e-05,
+      "loss": 1.4324,
+      "step": 6927
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8697865009307861,
+      "learning_rate": 4.8171292624900877e-05,
+      "loss": 1.5253,
+      "step": 6928
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2280068397521973,
+      "learning_rate": 4.817102828443035e-05,
+      "loss": 1.3758,
+      "step": 6929
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.786478281021118,
+      "learning_rate": 4.8170763943959824e-05,
+      "loss": 1.409,
+      "step": 6930
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0296168327331543,
+      "learning_rate": 4.81704996034893e-05,
+      "loss": 1.6019,
+      "step": 6931
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9051936864852905,
+      "learning_rate": 4.817023526301877e-05,
+      "loss": 1.4159,
+      "step": 6932
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6593403816223145,
+      "learning_rate": 4.8169970922548244e-05,
+      "loss": 1.3782,
+      "step": 6933
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5261547565460205,
+      "learning_rate": 4.816970658207772e-05,
+      "loss": 1.3886,
+      "step": 6934
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6689740419387817,
+      "learning_rate": 4.816944224160719e-05,
+      "loss": 1.3422,
+      "step": 6935
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5137028694152832,
+      "learning_rate": 4.8169177901136665e-05,
+      "loss": 1.3026,
+      "step": 6936
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4208792448043823,
+      "learning_rate": 4.816891356066614e-05,
+      "loss": 1.3078,
+      "step": 6937
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5884186029434204,
+      "learning_rate": 4.816864922019562e-05,
+      "loss": 1.2936,
+      "step": 6938
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.010342836380005,
+      "learning_rate": 4.8168384879725086e-05,
+      "loss": 1.4857,
+      "step": 6939
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1396470069885254,
+      "learning_rate": 4.8168120539254566e-05,
+      "loss": 1.5821,
+      "step": 6940
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0556843280792236,
+      "learning_rate": 4.816785619878403e-05,
+      "loss": 1.5012,
+      "step": 6941
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.584794282913208,
+      "learning_rate": 4.816759185831351e-05,
+      "loss": 1.7975,
+      "step": 6942
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.539850115776062,
+      "learning_rate": 4.816732751784299e-05,
+      "loss": 1.3431,
+      "step": 6943
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.778058409690857,
+      "learning_rate": 4.8167063177372454e-05,
+      "loss": 1.4014,
+      "step": 6944
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9547181129455566,
+      "learning_rate": 4.8166798836901934e-05,
+      "loss": 1.4445,
+      "step": 6945
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9877065420150757,
+      "learning_rate": 4.81665344964314e-05,
+      "loss": 1.5823,
+      "step": 6946
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7081060409545898,
+      "learning_rate": 4.816627015596088e-05,
+      "loss": 1.3983,
+      "step": 6947
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.149670124053955,
+      "learning_rate": 4.8166005815490355e-05,
+      "loss": 1.3381,
+      "step": 6948
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7267405986785889,
+      "learning_rate": 4.816574147501983e-05,
+      "loss": 1.4438,
+      "step": 6949
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.694516897201538,
+      "learning_rate": 4.81654771345493e-05,
+      "loss": 1.461,
+      "step": 6950
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.625123143196106,
+      "learning_rate": 4.8165212794078775e-05,
+      "loss": 1.366,
+      "step": 6951
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7835770845413208,
+      "learning_rate": 4.816494845360825e-05,
+      "loss": 1.3724,
+      "step": 6952
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9930158853530884,
+      "learning_rate": 4.816468411313772e-05,
+      "loss": 1.4386,
+      "step": 6953
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.113771915435791,
+      "learning_rate": 4.8164419772667196e-05,
+      "loss": 1.5406,
+      "step": 6954
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.402386426925659,
+      "learning_rate": 4.816415543219667e-05,
+      "loss": 1.4339,
+      "step": 6955
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7326699495315552,
+      "learning_rate": 4.816389109172614e-05,
+      "loss": 1.4821,
+      "step": 6956
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6677826642990112,
+      "learning_rate": 4.816362675125562e-05,
+      "loss": 1.472,
+      "step": 6957
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1624085903167725,
+      "learning_rate": 4.81633624107851e-05,
+      "loss": 1.4274,
+      "step": 6958
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.100282907485962,
+      "learning_rate": 4.8163098070314564e-05,
+      "loss": 1.5576,
+      "step": 6959
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0414702892303467,
+      "learning_rate": 4.8162833729844044e-05,
+      "loss": 1.2999,
+      "step": 6960
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4405814409255981,
+      "learning_rate": 4.816256938937351e-05,
+      "loss": 1.3742,
+      "step": 6961
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7186604738235474,
+      "learning_rate": 4.816230504890299e-05,
+      "loss": 1.3872,
+      "step": 6962
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7242158651351929,
+      "learning_rate": 4.8162040708432465e-05,
+      "loss": 1.1803,
+      "step": 6963
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8950825929641724,
+      "learning_rate": 4.816177636796194e-05,
+      "loss": 1.496,
+      "step": 6964
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4235045909881592,
+      "learning_rate": 4.816151202749141e-05,
+      "loss": 1.3546,
+      "step": 6965
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3284363746643066,
+      "learning_rate": 4.8161247687020886e-05,
+      "loss": 1.3379,
+      "step": 6966
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8885418176651,
+      "learning_rate": 4.816098334655036e-05,
+      "loss": 1.3359,
+      "step": 6967
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6479870080947876,
+      "learning_rate": 4.816071900607983e-05,
+      "loss": 1.3317,
+      "step": 6968
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.32211434841156,
+      "learning_rate": 4.8160454665609306e-05,
+      "loss": 2.4086,
+      "step": 6969
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3856931924819946,
+      "learning_rate": 4.816019032513878e-05,
+      "loss": 1.6563,
+      "step": 6970
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9347447156906128,
+      "learning_rate": 4.8159925984668254e-05,
+      "loss": 1.4083,
+      "step": 6971
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4938993453979492,
+      "learning_rate": 4.815966164419773e-05,
+      "loss": 1.4313,
+      "step": 6972
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.654768705368042,
+      "learning_rate": 4.815939730372721e-05,
+      "loss": 1.4886,
+      "step": 6973
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2993173599243164,
+      "learning_rate": 4.8159132963256674e-05,
+      "loss": 1.511,
+      "step": 6974
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6326541900634766,
+      "learning_rate": 4.8158868622786155e-05,
+      "loss": 1.3378,
+      "step": 6975
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.071922540664673,
+      "learning_rate": 4.815860428231562e-05,
+      "loss": 1.4562,
+      "step": 6976
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.488232135772705,
+      "learning_rate": 4.81583399418451e-05,
+      "loss": 1.5686,
+      "step": 6977
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.824216842651367,
+      "learning_rate": 4.8158075601374575e-05,
+      "loss": 1.4552,
+      "step": 6978
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8393018245697021,
+      "learning_rate": 4.815781126090404e-05,
+      "loss": 1.2856,
+      "step": 6979
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3837847709655762,
+      "learning_rate": 4.815754692043352e-05,
+      "loss": 1.3214,
+      "step": 6980
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9774078130722046,
+      "learning_rate": 4.815728257996299e-05,
+      "loss": 1.6775,
+      "step": 6981
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8646010160446167,
+      "learning_rate": 4.815701823949247e-05,
+      "loss": 1.4663,
+      "step": 6982
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5773775577545166,
+      "learning_rate": 4.815675389902194e-05,
+      "loss": 1.4312,
+      "step": 6983
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.034886121749878,
+      "learning_rate": 4.815648955855142e-05,
+      "loss": 1.4709,
+      "step": 6984
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.78562331199646,
+      "learning_rate": 4.815622521808089e-05,
+      "loss": 1.5185,
+      "step": 6985
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5444977283477783,
+      "learning_rate": 4.8155960877610364e-05,
+      "loss": 1.2521,
+      "step": 6986
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4263203144073486,
+      "learning_rate": 4.815569653713984e-05,
+      "loss": 1.2379,
+      "step": 6987
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5067243576049805,
+      "learning_rate": 4.815543219666931e-05,
+      "loss": 1.1261,
+      "step": 6988
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4359166622161865,
+      "learning_rate": 4.8155167856198785e-05,
+      "loss": 1.2637,
+      "step": 6989
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.810145616531372,
+      "learning_rate": 4.815490351572826e-05,
+      "loss": 1.3928,
+      "step": 6990
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8045846223831177,
+      "learning_rate": 4.815463917525773e-05,
+      "loss": 1.4883,
+      "step": 6991
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6191020011901855,
+      "learning_rate": 4.8154374834787205e-05,
+      "loss": 1.4105,
+      "step": 6992
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.034217596054077,
+      "learning_rate": 4.8154110494316686e-05,
+      "loss": 1.3935,
+      "step": 6993
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.904018759727478,
+      "learning_rate": 4.815384615384615e-05,
+      "loss": 1.5228,
+      "step": 6994
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.904375433921814,
+      "learning_rate": 4.815358181337563e-05,
+      "loss": 1.3601,
+      "step": 6995
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8782974481582642,
+      "learning_rate": 4.81533174729051e-05,
+      "loss": 1.442,
+      "step": 6996
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.041640281677246,
+      "learning_rate": 4.815305313243458e-05,
+      "loss": 1.528,
+      "step": 6997
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0045440196990967,
+      "learning_rate": 4.8152788791964054e-05,
+      "loss": 2.2481,
+      "step": 6998
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.715049982070923,
+      "learning_rate": 4.815252445149353e-05,
+      "loss": 1.6195,
+      "step": 6999
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9878873825073242,
+      "learning_rate": 4.8152260111023e-05,
+      "loss": 1.4919,
+      "step": 7000
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.559253215789795,
+      "learning_rate": 4.8151995770552474e-05,
+      "loss": 1.3188,
+      "step": 7001
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3064942359924316,
+      "learning_rate": 4.815173143008195e-05,
+      "loss": 1.4163,
+      "step": 7002
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5596985816955566,
+      "learning_rate": 4.815146708961142e-05,
+      "loss": 1.3144,
+      "step": 7003
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2272613048553467,
+      "learning_rate": 4.8151202749140895e-05,
+      "loss": 1.4935,
+      "step": 7004
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9846258163452148,
+      "learning_rate": 4.815093840867037e-05,
+      "loss": 1.5859,
+      "step": 7005
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.316420316696167,
+      "learning_rate": 4.815067406819984e-05,
+      "loss": 1.4066,
+      "step": 7006
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7884949445724487,
+      "learning_rate": 4.8150409727729316e-05,
+      "loss": 1.4719,
+      "step": 7007
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.656618595123291,
+      "learning_rate": 4.8150145387258796e-05,
+      "loss": 1.4089,
+      "step": 7008
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.708837628364563,
+      "learning_rate": 4.814988104678826e-05,
+      "loss": 1.3642,
+      "step": 7009
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.138536214828491,
+      "learning_rate": 4.814961670631774e-05,
+      "loss": 1.4901,
+      "step": 7010
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0971381664276123,
+      "learning_rate": 4.814935236584721e-05,
+      "loss": 2.5398,
+      "step": 7011
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0382070541381836,
+      "learning_rate": 4.814908802537669e-05,
+      "loss": 1.4759,
+      "step": 7012
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4423296451568604,
+      "learning_rate": 4.8148823684906164e-05,
+      "loss": 1.3664,
+      "step": 7013
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8613500595092773,
+      "learning_rate": 4.814855934443564e-05,
+      "loss": 1.3239,
+      "step": 7014
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6409708261489868,
+      "learning_rate": 4.814829500396511e-05,
+      "loss": 1.3169,
+      "step": 7015
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5381877422332764,
+      "learning_rate": 4.814803066349458e-05,
+      "loss": 1.8637,
+      "step": 7016
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2802726030349731,
+      "learning_rate": 4.814776632302406e-05,
+      "loss": 1.1129,
+      "step": 7017
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.500720500946045,
+      "learning_rate": 4.814750198255353e-05,
+      "loss": 2.1349,
+      "step": 7018
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9639668464660645,
+      "learning_rate": 4.8147237642083005e-05,
+      "loss": 1.5496,
+      "step": 7019
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8806639909744263,
+      "learning_rate": 4.814697330161248e-05,
+      "loss": 1.394,
+      "step": 7020
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0583367347717285,
+      "learning_rate": 4.814670896114195e-05,
+      "loss": 1.2799,
+      "step": 7021
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7149044275283813,
+      "learning_rate": 4.8146444620671426e-05,
+      "loss": 1.4217,
+      "step": 7022
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6879464387893677,
+      "learning_rate": 4.8146180280200906e-05,
+      "loss": 1.4095,
+      "step": 7023
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6256722211837769,
+      "learning_rate": 4.814591593973037e-05,
+      "loss": 1.2831,
+      "step": 7024
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.143699884414673,
+      "learning_rate": 4.814565159925985e-05,
+      "loss": 1.4099,
+      "step": 7025
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.896765947341919,
+      "learning_rate": 4.814538725878932e-05,
+      "loss": 1.2975,
+      "step": 7026
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9173405170440674,
+      "learning_rate": 4.8145122918318794e-05,
+      "loss": 1.3879,
+      "step": 7027
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3416342735290527,
+      "learning_rate": 4.8144858577848274e-05,
+      "loss": 1.4869,
+      "step": 7028
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9594484567642212,
+      "learning_rate": 4.814459423737774e-05,
+      "loss": 1.4163,
+      "step": 7029
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8975247144699097,
+      "learning_rate": 4.814432989690722e-05,
+      "loss": 1.4694,
+      "step": 7030
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.544425368309021,
+      "learning_rate": 4.814406555643669e-05,
+      "loss": 1.357,
+      "step": 7031
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0792994499206543,
+      "learning_rate": 4.814380121596617e-05,
+      "loss": 1.3458,
+      "step": 7032
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4999090433120728,
+      "learning_rate": 4.814353687549564e-05,
+      "loss": 1.374,
+      "step": 7033
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1653616428375244,
+      "learning_rate": 4.8143272535025116e-05,
+      "loss": 1.4844,
+      "step": 7034
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8895316123962402,
+      "learning_rate": 4.814300819455459e-05,
+      "loss": 1.2548,
+      "step": 7035
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2592973709106445,
+      "learning_rate": 4.814274385408406e-05,
+      "loss": 1.4345,
+      "step": 7036
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.072275400161743,
+      "learning_rate": 4.8142479513613536e-05,
+      "loss": 1.5407,
+      "step": 7037
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7087512016296387,
+      "learning_rate": 4.814221517314301e-05,
+      "loss": 1.4602,
+      "step": 7038
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6337134838104248,
+      "learning_rate": 4.8141950832672484e-05,
+      "loss": 1.2644,
+      "step": 7039
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9092127084732056,
+      "learning_rate": 4.814168649220196e-05,
+      "loss": 1.3771,
+      "step": 7040
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5727862119674683,
+      "learning_rate": 4.814142215173143e-05,
+      "loss": 1.4326,
+      "step": 7041
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7609829902648926,
+      "learning_rate": 4.8141157811260904e-05,
+      "loss": 1.4365,
+      "step": 7042
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9814705848693848,
+      "learning_rate": 4.8140893470790385e-05,
+      "loss": 1.4629,
+      "step": 7043
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.543951153755188,
+      "learning_rate": 4.814062913031985e-05,
+      "loss": 1.3323,
+      "step": 7044
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0717413425445557,
+      "learning_rate": 4.814036478984933e-05,
+      "loss": 1.5121,
+      "step": 7045
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5845292806625366,
+      "learning_rate": 4.81401004493788e-05,
+      "loss": 1.3943,
+      "step": 7046
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5930787324905396,
+      "learning_rate": 4.813983610890828e-05,
+      "loss": 1.3947,
+      "step": 7047
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8851876258850098,
+      "learning_rate": 4.813957176843775e-05,
+      "loss": 1.421,
+      "step": 7048
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5995688438415527,
+      "learning_rate": 4.8139307427967226e-05,
+      "loss": 1.5121,
+      "step": 7049
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.368864059448242,
+      "learning_rate": 4.81390430874967e-05,
+      "loss": 1.5647,
+      "step": 7050
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8710423707962036,
+      "learning_rate": 4.8138778747026166e-05,
+      "loss": 1.3747,
+      "step": 7051
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3803514242172241,
+      "learning_rate": 4.813851440655565e-05,
+      "loss": 1.3255,
+      "step": 7052
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4615495204925537,
+      "learning_rate": 4.813825006608512e-05,
+      "loss": 1.3401,
+      "step": 7053
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.646640658378601,
+      "learning_rate": 4.8137985725614594e-05,
+      "loss": 1.3695,
+      "step": 7054
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8055402040481567,
+      "learning_rate": 4.813772138514407e-05,
+      "loss": 1.4524,
+      "step": 7055
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7384485006332397,
+      "learning_rate": 4.813745704467354e-05,
+      "loss": 1.3539,
+      "step": 7056
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.744855284690857,
+      "learning_rate": 4.8137192704203015e-05,
+      "loss": 1.42,
+      "step": 7057
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8813824653625488,
+      "learning_rate": 4.8136928363732495e-05,
+      "loss": 1.409,
+      "step": 7058
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8783966302871704,
+      "learning_rate": 4.813666402326196e-05,
+      "loss": 1.3943,
+      "step": 7059
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6356078386306763,
+      "learning_rate": 4.813639968279144e-05,
+      "loss": 1.3679,
+      "step": 7060
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7657113075256348,
+      "learning_rate": 4.813613534232091e-05,
+      "loss": 1.4598,
+      "step": 7061
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.663906216621399,
+      "learning_rate": 4.813587100185038e-05,
+      "loss": 1.355,
+      "step": 7062
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7684825658798218,
+      "learning_rate": 4.813560666137986e-05,
+      "loss": 1.4919,
+      "step": 7063
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9655641317367554,
+      "learning_rate": 4.813534232090933e-05,
+      "loss": 1.5124,
+      "step": 7064
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5682979822158813,
+      "learning_rate": 4.813507798043881e-05,
+      "loss": 1.4052,
+      "step": 7065
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9500389099121094,
+      "learning_rate": 4.813481363996828e-05,
+      "loss": 1.3714,
+      "step": 7066
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.13362717628479,
+      "learning_rate": 4.813454929949776e-05,
+      "loss": 1.4158,
+      "step": 7067
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.57034170627594,
+      "learning_rate": 4.813428495902723e-05,
+      "loss": 1.4491,
+      "step": 7068
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.643128752708435,
+      "learning_rate": 4.8134020618556704e-05,
+      "loss": 1.2954,
+      "step": 7069
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1251235008239746,
+      "learning_rate": 4.813375627808618e-05,
+      "loss": 1.5639,
+      "step": 7070
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.730125069618225,
+      "learning_rate": 4.813349193761565e-05,
+      "loss": 1.3267,
+      "step": 7071
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9166553020477295,
+      "learning_rate": 4.8133227597145125e-05,
+      "loss": 1.331,
+      "step": 7072
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6172679662704468,
+      "learning_rate": 4.81329632566746e-05,
+      "loss": 1.2538,
+      "step": 7073
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5673733949661255,
+      "learning_rate": 4.813269891620407e-05,
+      "loss": 1.2742,
+      "step": 7074
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2261364459991455,
+      "learning_rate": 4.8132434575733546e-05,
+      "loss": 1.5559,
+      "step": 7075
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6006006002426147,
+      "learning_rate": 4.813217023526302e-05,
+      "loss": 1.4274,
+      "step": 7076
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6745401620864868,
+      "learning_rate": 4.813190589479249e-05,
+      "loss": 1.4512,
+      "step": 7077
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0173494815826416,
+      "learning_rate": 4.813164155432197e-05,
+      "loss": 1.4358,
+      "step": 7078
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.076082944869995,
+      "learning_rate": 4.813137721385144e-05,
+      "loss": 1.5706,
+      "step": 7079
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.028641939163208,
+      "learning_rate": 4.813111287338092e-05,
+      "loss": 1.4646,
+      "step": 7080
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0810186862945557,
+      "learning_rate": 4.813084853291039e-05,
+      "loss": 1.4472,
+      "step": 7081
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9018144607543945,
+      "learning_rate": 4.813058419243987e-05,
+      "loss": 1.5611,
+      "step": 7082
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.515683889389038,
+      "learning_rate": 4.813031985196934e-05,
+      "loss": 1.3752,
+      "step": 7083
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.863921046257019,
+      "learning_rate": 4.8130055511498815e-05,
+      "loss": 1.5682,
+      "step": 7084
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9596247673034668,
+      "learning_rate": 4.812979117102829e-05,
+      "loss": 1.4981,
+      "step": 7085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.615418553352356,
+      "learning_rate": 4.812952683055776e-05,
+      "loss": 1.3793,
+      "step": 7086
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.655488133430481,
+      "learning_rate": 4.8129262490087235e-05,
+      "loss": 1.4655,
+      "step": 7087
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4563162326812744,
+      "learning_rate": 4.812899814961671e-05,
+      "loss": 1.4663,
+      "step": 7088
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6432496309280396,
+      "learning_rate": 4.812873380914618e-05,
+      "loss": 1.529,
+      "step": 7089
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5568857192993164,
+      "learning_rate": 4.8128469468675656e-05,
+      "loss": 1.351,
+      "step": 7090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8601876497268677,
+      "learning_rate": 4.812820512820513e-05,
+      "loss": 1.3941,
+      "step": 7091
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5864262580871582,
+      "learning_rate": 4.81279407877346e-05,
+      "loss": 1.4076,
+      "step": 7092
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3310585021972656,
+      "learning_rate": 4.8127676447264084e-05,
+      "loss": 1.408,
+      "step": 7093
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.527449369430542,
+      "learning_rate": 4.812741210679355e-05,
+      "loss": 1.3911,
+      "step": 7094
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7184627056121826,
+      "learning_rate": 4.812714776632303e-05,
+      "loss": 1.3927,
+      "step": 7095
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.42533540725708,
+      "learning_rate": 4.81268834258525e-05,
+      "loss": 1.5569,
+      "step": 7096
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5899337530136108,
+      "learning_rate": 4.812661908538197e-05,
+      "loss": 1.3086,
+      "step": 7097
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2799899578094482,
+      "learning_rate": 4.812635474491145e-05,
+      "loss": 1.4558,
+      "step": 7098
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5568169355392456,
+      "learning_rate": 4.812609040444092e-05,
+      "loss": 1.2414,
+      "step": 7099
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.609317660331726,
+      "learning_rate": 4.81258260639704e-05,
+      "loss": 1.2081,
+      "step": 7100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4846792221069336,
+      "learning_rate": 4.8125561723499865e-05,
+      "loss": 1.433,
+      "step": 7101
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6482990980148315,
+      "learning_rate": 4.8125297383029346e-05,
+      "loss": 1.406,
+      "step": 7102
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6056602001190186,
+      "learning_rate": 4.812503304255882e-05,
+      "loss": 1.3313,
+      "step": 7103
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6431964635849,
+      "learning_rate": 4.812476870208829e-05,
+      "loss": 1.5083,
+      "step": 7104
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6153366565704346,
+      "learning_rate": 4.8124504361617766e-05,
+      "loss": 1.6411,
+      "step": 7105
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8151969909667969,
+      "learning_rate": 4.812424002114724e-05,
+      "loss": 1.4834,
+      "step": 7106
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7918139696121216,
+      "learning_rate": 4.8123975680676713e-05,
+      "loss": 1.4838,
+      "step": 7107
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.518136978149414,
+      "learning_rate": 4.812371134020619e-05,
+      "loss": 1.2959,
+      "step": 7108
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.050840139389038,
+      "learning_rate": 4.812344699973566e-05,
+      "loss": 1.2111,
+      "step": 7109
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4236507415771484,
+      "learning_rate": 4.8123182659265134e-05,
+      "loss": 1.2097,
+      "step": 7110
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5491969585418701,
+      "learning_rate": 4.812291831879461e-05,
+      "loss": 1.3408,
+      "step": 7111
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3058823347091675,
+      "learning_rate": 4.812265397832408e-05,
+      "loss": 1.2465,
+      "step": 7112
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6627235412597656,
+      "learning_rate": 4.812238963785356e-05,
+      "loss": 1.3038,
+      "step": 7113
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9627950191497803,
+      "learning_rate": 4.812212529738303e-05,
+      "loss": 1.5452,
+      "step": 7114
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.86781907081604,
+      "learning_rate": 4.812186095691251e-05,
+      "loss": 1.4556,
+      "step": 7115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8184478282928467,
+      "learning_rate": 4.8121596616441976e-05,
+      "loss": 1.5598,
+      "step": 7116
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.936367154121399,
+      "learning_rate": 4.8121332275971456e-05,
+      "loss": 2.1223,
+      "step": 7117
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3756535053253174,
+      "learning_rate": 4.812106793550093e-05,
+      "loss": 1.079,
+      "step": 7118
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8044989109039307,
+      "learning_rate": 4.81208035950304e-05,
+      "loss": 1.3168,
+      "step": 7119
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.18527889251709,
+      "learning_rate": 4.812053925455988e-05,
+      "loss": 1.467,
+      "step": 7120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7225972414016724,
+      "learning_rate": 4.812027491408935e-05,
+      "loss": 1.4136,
+      "step": 7121
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0006141662597656,
+      "learning_rate": 4.8120010573618824e-05,
+      "loss": 1.5429,
+      "step": 7122
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5557504892349243,
+      "learning_rate": 4.81197462331483e-05,
+      "loss": 2.8704,
+      "step": 7123
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7546539306640625,
+      "learning_rate": 4.811948189267777e-05,
+      "loss": 1.5425,
+      "step": 7124
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8838999271392822,
+      "learning_rate": 4.8119217552207245e-05,
+      "loss": 1.5669,
+      "step": 7125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9905753135681152,
+      "learning_rate": 4.811895321173672e-05,
+      "loss": 1.5294,
+      "step": 7126
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1209888458251953,
+      "learning_rate": 4.811868887126619e-05,
+      "loss": 1.5577,
+      "step": 7127
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6233320236206055,
+      "learning_rate": 4.811842453079567e-05,
+      "loss": 1.3608,
+      "step": 7128
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7205698490142822,
+      "learning_rate": 4.811816019032514e-05,
+      "loss": 1.486,
+      "step": 7129
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7260891199111938,
+      "learning_rate": 4.811789584985462e-05,
+      "loss": 1.484,
+      "step": 7130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8614078760147095,
+      "learning_rate": 4.8117631509384086e-05,
+      "loss": 1.3451,
+      "step": 7131
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4476031064987183,
+      "learning_rate": 4.8117367168913566e-05,
+      "loss": 2.0356,
+      "step": 7132
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7250722646713257,
+      "learning_rate": 4.811710282844304e-05,
+      "loss": 1.4081,
+      "step": 7133
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.670236825942993,
+      "learning_rate": 4.811683848797251e-05,
+      "loss": 1.527,
+      "step": 7134
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3227274417877197,
+      "learning_rate": 4.811657414750199e-05,
+      "loss": 1.4129,
+      "step": 7135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.157978057861328,
+      "learning_rate": 4.8116309807031454e-05,
+      "loss": 1.3661,
+      "step": 7136
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.565036654472351,
+      "learning_rate": 4.8116045466560934e-05,
+      "loss": 1.4513,
+      "step": 7137
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.84847354888916,
+      "learning_rate": 4.811578112609041e-05,
+      "loss": 1.437,
+      "step": 7138
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.273650884628296,
+      "learning_rate": 4.811551678561988e-05,
+      "loss": 1.4254,
+      "step": 7139
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.132603406906128,
+      "learning_rate": 4.8115252445149355e-05,
+      "loss": 1.4491,
+      "step": 7140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0607051849365234,
+      "learning_rate": 4.811498810467883e-05,
+      "loss": 1.4708,
+      "step": 7141
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3763338327407837,
+      "learning_rate": 4.81147237642083e-05,
+      "loss": 1.2533,
+      "step": 7142
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7751097679138184,
+      "learning_rate": 4.8114459423737776e-05,
+      "loss": 1.2731,
+      "step": 7143
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0937228202819824,
+      "learning_rate": 4.811419508326725e-05,
+      "loss": 1.4631,
+      "step": 7144
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9779746532440186,
+      "learning_rate": 4.811393074279672e-05,
+      "loss": 1.3648,
+      "step": 7145
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6087086200714111,
+      "learning_rate": 4.8113666402326196e-05,
+      "loss": 1.2331,
+      "step": 7146
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9666826725006104,
+      "learning_rate": 4.811340206185567e-05,
+      "loss": 1.3844,
+      "step": 7147
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4773645401000977,
+      "learning_rate": 4.811313772138515e-05,
+      "loss": 1.5875,
+      "step": 7148
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.232131004333496,
+      "learning_rate": 4.811287338091462e-05,
+      "loss": 1.8717,
+      "step": 7149
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7883949279785156,
+      "learning_rate": 4.81126090404441e-05,
+      "loss": 1.4655,
+      "step": 7150
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5653259754180908,
+      "learning_rate": 4.8112344699973564e-05,
+      "loss": 1.2567,
+      "step": 7151
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9676988124847412,
+      "learning_rate": 4.8112080359503045e-05,
+      "loss": 1.3942,
+      "step": 7152
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0017590522766113,
+      "learning_rate": 4.811181601903252e-05,
+      "loss": 1.5569,
+      "step": 7153
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1515557765960693,
+      "learning_rate": 4.811155167856199e-05,
+      "loss": 1.4347,
+      "step": 7154
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8179829120635986,
+      "learning_rate": 4.8111287338091465e-05,
+      "loss": 1.5407,
+      "step": 7155
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.716442584991455,
+      "learning_rate": 4.811102299762094e-05,
+      "loss": 1.2403,
+      "step": 7156
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4412096738815308,
+      "learning_rate": 4.811075865715041e-05,
+      "loss": 1.3577,
+      "step": 7157
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9305014610290527,
+      "learning_rate": 4.8110494316679886e-05,
+      "loss": 1.3722,
+      "step": 7158
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4751302003860474,
+      "learning_rate": 4.811022997620936e-05,
+      "loss": 1.2042,
+      "step": 7159
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.078244209289551,
+      "learning_rate": 4.810996563573883e-05,
+      "loss": 1.5534,
+      "step": 7160
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0399653911590576,
+      "learning_rate": 4.810970129526831e-05,
+      "loss": 1.4336,
+      "step": 7161
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.893659234046936,
+      "learning_rate": 4.810943695479778e-05,
+      "loss": 1.4604,
+      "step": 7162
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4650073051452637,
+      "learning_rate": 4.810917261432726e-05,
+      "loss": 1.1012,
+      "step": 7163
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5510404109954834,
+      "learning_rate": 4.810890827385673e-05,
+      "loss": 1.3419,
+      "step": 7164
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.707729697227478,
+      "learning_rate": 4.810864393338621e-05,
+      "loss": 1.3399,
+      "step": 7165
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3620052337646484,
+      "learning_rate": 4.8108379592915675e-05,
+      "loss": 1.6196,
+      "step": 7166
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.700670599937439,
+      "learning_rate": 4.8108115252445155e-05,
+      "loss": 2.5944,
+      "step": 7167
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.65610933303833,
+      "learning_rate": 4.810785091197463e-05,
+      "loss": 1.2516,
+      "step": 7168
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8163390159606934,
+      "learning_rate": 4.8107586571504095e-05,
+      "loss": 1.4255,
+      "step": 7169
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.611061692237854,
+      "learning_rate": 4.8107322231033576e-05,
+      "loss": 1.2911,
+      "step": 7170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.693332552909851,
+      "learning_rate": 4.810705789056304e-05,
+      "loss": 1.6318,
+      "step": 7171
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.267737865447998,
+      "learning_rate": 4.810679355009252e-05,
+      "loss": 1.5223,
+      "step": 7172
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9291622638702393,
+      "learning_rate": 4.8106529209621996e-05,
+      "loss": 1.3725,
+      "step": 7173
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9228436946868896,
+      "learning_rate": 4.810626486915147e-05,
+      "loss": 2.4678,
+      "step": 7174
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.656543016433716,
+      "learning_rate": 4.8106000528680943e-05,
+      "loss": 1.5284,
+      "step": 7175
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5021624565124512,
+      "learning_rate": 4.810573618821042e-05,
+      "loss": 1.3412,
+      "step": 7176
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.514814019203186,
+      "learning_rate": 4.810547184773989e-05,
+      "loss": 1.4389,
+      "step": 7177
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.405019998550415,
+      "learning_rate": 4.810520750726937e-05,
+      "loss": 1.5212,
+      "step": 7178
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7211240530014038,
+      "learning_rate": 4.810494316679884e-05,
+      "loss": 1.5413,
+      "step": 7179
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8959362506866455,
+      "learning_rate": 4.810467882632831e-05,
+      "loss": 1.3103,
+      "step": 7180
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7430505752563477,
+      "learning_rate": 4.8104414485857785e-05,
+      "loss": 1.4912,
+      "step": 7181
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.513283610343933,
+      "learning_rate": 4.810415014538726e-05,
+      "loss": 1.2951,
+      "step": 7182
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.221017837524414,
+      "learning_rate": 4.810388580491674e-05,
+      "loss": 1.5888,
+      "step": 7183
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.40781569480896,
+      "learning_rate": 4.8103621464446206e-05,
+      "loss": 1.1837,
+      "step": 7184
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2785627841949463,
+      "learning_rate": 4.8103357123975686e-05,
+      "loss": 1.343,
+      "step": 7185
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7459434270858765,
+      "learning_rate": 4.810309278350515e-05,
+      "loss": 1.3393,
+      "step": 7186
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7394598722457886,
+      "learning_rate": 4.810282844303463e-05,
+      "loss": 1.3096,
+      "step": 7187
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4122817516326904,
+      "learning_rate": 4.810256410256411e-05,
+      "loss": 1.3706,
+      "step": 7188
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7827030420303345,
+      "learning_rate": 4.810229976209358e-05,
+      "loss": 1.4152,
+      "step": 7189
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9859776496887207,
+      "learning_rate": 4.8102035421623054e-05,
+      "loss": 1.352,
+      "step": 7190
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0253865718841553,
+      "learning_rate": 4.810177108115253e-05,
+      "loss": 1.4245,
+      "step": 7191
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.463814616203308,
+      "learning_rate": 4.8101506740682e-05,
+      "loss": 1.2685,
+      "step": 7192
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.088254928588867,
+      "learning_rate": 4.8101242400211474e-05,
+      "loss": 1.353,
+      "step": 7193
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2390220165252686,
+      "learning_rate": 4.810097805974095e-05,
+      "loss": 1.4892,
+      "step": 7194
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8194589614868164,
+      "learning_rate": 4.810071371927042e-05,
+      "loss": 1.452,
+      "step": 7195
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.828076720237732,
+      "learning_rate": 4.8100449378799895e-05,
+      "loss": 1.4133,
+      "step": 7196
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5530925989151,
+      "learning_rate": 4.810018503832937e-05,
+      "loss": 1.2678,
+      "step": 7197
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4455093145370483,
+      "learning_rate": 4.809992069785885e-05,
+      "loss": 1.3923,
+      "step": 7198
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8661015033721924,
+      "learning_rate": 4.8099656357388316e-05,
+      "loss": 1.3394,
+      "step": 7199
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0011403560638428,
+      "learning_rate": 4.8099392016917796e-05,
+      "loss": 1.2927,
+      "step": 7200
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7838540077209473,
+      "learning_rate": 4.809912767644726e-05,
+      "loss": 1.3904,
+      "step": 7201
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7170228958129883,
+      "learning_rate": 4.8098863335976743e-05,
+      "loss": 1.3087,
+      "step": 7202
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9612151384353638,
+      "learning_rate": 4.809859899550622e-05,
+      "loss": 1.4034,
+      "step": 7203
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7967467308044434,
+      "learning_rate": 4.809833465503569e-05,
+      "loss": 1.3965,
+      "step": 7204
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.10269832611084,
+      "learning_rate": 4.8098070314565164e-05,
+      "loss": 1.4557,
+      "step": 7205
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.23891544342041,
+      "learning_rate": 4.809780597409463e-05,
+      "loss": 1.4028,
+      "step": 7206
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.340328574180603,
+      "learning_rate": 4.809754163362411e-05,
+      "loss": 1.1264,
+      "step": 7207
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5024220943450928,
+      "learning_rate": 4.8097277293153585e-05,
+      "loss": 1.6997,
+      "step": 7208
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0387086868286133,
+      "learning_rate": 4.809701295268306e-05,
+      "loss": 1.5458,
+      "step": 7209
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8047986030578613,
+      "learning_rate": 4.809674861221253e-05,
+      "loss": 1.3929,
+      "step": 7210
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8537780046463013,
+      "learning_rate": 4.8096484271742006e-05,
+      "loss": 1.5244,
+      "step": 7211
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0754761695861816,
+      "learning_rate": 4.809621993127148e-05,
+      "loss": 1.335,
+      "step": 7212
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.083516836166382,
+      "learning_rate": 4.809595559080096e-05,
+      "loss": 1.4524,
+      "step": 7213
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5959084033966064,
+      "learning_rate": 4.8095691250330426e-05,
+      "loss": 1.2865,
+      "step": 7214
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6993831396102905,
+      "learning_rate": 4.80954269098599e-05,
+      "loss": 1.4224,
+      "step": 7215
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7348167896270752,
+      "learning_rate": 4.809516256938937e-05,
+      "loss": 1.3857,
+      "step": 7216
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.685356616973877,
+      "learning_rate": 4.809489822891885e-05,
+      "loss": 1.3098,
+      "step": 7217
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1100709438323975,
+      "learning_rate": 4.809463388844833e-05,
+      "loss": 1.533,
+      "step": 7218
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6206474304199219,
+      "learning_rate": 4.8094369547977794e-05,
+      "loss": 1.3174,
+      "step": 7219
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1682989597320557,
+      "learning_rate": 4.8094105207507274e-05,
+      "loss": 1.4825,
+      "step": 7220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1195850372314453,
+      "learning_rate": 4.809384086703674e-05,
+      "loss": 1.457,
+      "step": 7221
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.100210428237915,
+      "learning_rate": 4.809357652656622e-05,
+      "loss": 1.5194,
+      "step": 7222
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6847819089889526,
+      "learning_rate": 4.8093312186095695e-05,
+      "loss": 1.4685,
+      "step": 7223
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7572754621505737,
+      "learning_rate": 4.809304784562517e-05,
+      "loss": 1.4318,
+      "step": 7224
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6530555486679077,
+      "learning_rate": 4.809278350515464e-05,
+      "loss": 1.1892,
+      "step": 7225
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2750775814056396,
+      "learning_rate": 4.8092519164684116e-05,
+      "loss": 1.5429,
+      "step": 7226
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.474229335784912,
+      "learning_rate": 4.809225482421359e-05,
+      "loss": 2.7422,
+      "step": 7227
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.860334038734436,
+      "learning_rate": 4.809199048374306e-05,
+      "loss": 1.5479,
+      "step": 7228
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0996921062469482,
+      "learning_rate": 4.8091726143272537e-05,
+      "loss": 1.5073,
+      "step": 7229
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7616060972213745,
+      "learning_rate": 4.809146180280201e-05,
+      "loss": 1.3712,
+      "step": 7230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7181363105773926,
+      "learning_rate": 4.8091197462331484e-05,
+      "loss": 1.5007,
+      "step": 7231
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0021817684173584,
+      "learning_rate": 4.809093312186096e-05,
+      "loss": 1.4268,
+      "step": 7232
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5130772590637207,
+      "learning_rate": 4.809066878139044e-05,
+      "loss": 1.3728,
+      "step": 7233
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5024957656860352,
+      "learning_rate": 4.8090404440919904e-05,
+      "loss": 1.266,
+      "step": 7234
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5332190990447998,
+      "learning_rate": 4.8090140100449385e-05,
+      "loss": 1.2531,
+      "step": 7235
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8440842628479004,
+      "learning_rate": 4.808987575997885e-05,
+      "loss": 1.2645,
+      "step": 7236
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.754975438117981,
+      "learning_rate": 4.808961141950833e-05,
+      "loss": 1.496,
+      "step": 7237
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9392775297164917,
+      "learning_rate": 4.8089347079037805e-05,
+      "loss": 1.4987,
+      "step": 7238
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1452009677886963,
+      "learning_rate": 4.808908273856728e-05,
+      "loss": 1.3369,
+      "step": 7239
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6066408157348633,
+      "learning_rate": 4.808881839809675e-05,
+      "loss": 1.6532,
+      "step": 7240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6153285503387451,
+      "learning_rate": 4.808855405762622e-05,
+      "loss": 1.3923,
+      "step": 7241
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8558658361434937,
+      "learning_rate": 4.80882897171557e-05,
+      "loss": 1.4801,
+      "step": 7242
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9519460201263428,
+      "learning_rate": 4.808802537668517e-05,
+      "loss": 1.3555,
+      "step": 7243
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7586203813552856,
+      "learning_rate": 4.808776103621465e-05,
+      "loss": 1.4787,
+      "step": 7244
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0816779136657715,
+      "learning_rate": 4.808749669574412e-05,
+      "loss": 1.6091,
+      "step": 7245
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.375110626220703,
+      "learning_rate": 4.8087232355273594e-05,
+      "loss": 1.3047,
+      "step": 7246
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9279201030731201,
+      "learning_rate": 4.808696801480307e-05,
+      "loss": 2.7694,
+      "step": 7247
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9567110538482666,
+      "learning_rate": 4.808670367433255e-05,
+      "loss": 1.3832,
+      "step": 7248
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5290260314941406,
+      "learning_rate": 4.8086439333862015e-05,
+      "loss": 1.3601,
+      "step": 7249
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5637646913528442,
+      "learning_rate": 4.8086174993391495e-05,
+      "loss": 1.4155,
+      "step": 7250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6455914974212646,
+      "learning_rate": 4.808591065292096e-05,
+      "loss": 1.3255,
+      "step": 7251
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7852258682250977,
+      "learning_rate": 4.8085646312450435e-05,
+      "loss": 1.4589,
+      "step": 7252
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.182084560394287,
+      "learning_rate": 4.8085381971979916e-05,
+      "loss": 1.6315,
+      "step": 7253
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.132568836212158,
+      "learning_rate": 4.808511763150938e-05,
+      "loss": 1.3741,
+      "step": 7254
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0921335220336914,
+      "learning_rate": 4.808485329103886e-05,
+      "loss": 2.378,
+      "step": 7255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.726008415222168,
+      "learning_rate": 4.808458895056833e-05,
+      "loss": 1.4125,
+      "step": 7256
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.501955270767212,
+      "learning_rate": 4.808432461009781e-05,
+      "loss": 1.3652,
+      "step": 7257
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8061273097991943,
+      "learning_rate": 4.8084060269627284e-05,
+      "loss": 1.3924,
+      "step": 7258
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6137720346450806,
+      "learning_rate": 4.808379592915676e-05,
+      "loss": 1.1262,
+      "step": 7259
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6808875799179077,
+      "learning_rate": 4.808353158868623e-05,
+      "loss": 1.2999,
+      "step": 7260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0923404693603516,
+      "learning_rate": 4.8083267248215704e-05,
+      "loss": 1.4315,
+      "step": 7261
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6104568243026733,
+      "learning_rate": 4.808300290774518e-05,
+      "loss": 1.4346,
+      "step": 7262
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2535055875778198,
+      "learning_rate": 4.808273856727465e-05,
+      "loss": 1.2894,
+      "step": 7263
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.478729724884033,
+      "learning_rate": 4.8082474226804125e-05,
+      "loss": 1.4118,
+      "step": 7264
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5648826360702515,
+      "learning_rate": 4.80822098863336e-05,
+      "loss": 1.3104,
+      "step": 7265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5969831943511963,
+      "learning_rate": 4.808194554586307e-05,
+      "loss": 1.4949,
+      "step": 7266
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9973043203353882,
+      "learning_rate": 4.8081681205392546e-05,
+      "loss": 1.3313,
+      "step": 7267
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9926165342330933,
+      "learning_rate": 4.8081416864922026e-05,
+      "loss": 1.3774,
+      "step": 7268
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9325836896896362,
+      "learning_rate": 4.808115252445149e-05,
+      "loss": 1.5582,
+      "step": 7269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5667433738708496,
+      "learning_rate": 4.808088818398097e-05,
+      "loss": 1.26,
+      "step": 7270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9760481119155884,
+      "learning_rate": 4.808062384351044e-05,
+      "loss": 1.3913,
+      "step": 7271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5775398015975952,
+      "learning_rate": 4.808035950303992e-05,
+      "loss": 1.3785,
+      "step": 7272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.248720407485962,
+      "learning_rate": 4.8080095162569394e-05,
+      "loss": 1.4483,
+      "step": 7273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.690799355506897,
+      "learning_rate": 4.807983082209887e-05,
+      "loss": 1.3442,
+      "step": 7274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6090179681777954,
+      "learning_rate": 4.807956648162834e-05,
+      "loss": 1.3129,
+      "step": 7275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8639923334121704,
+      "learning_rate": 4.8079302141157815e-05,
+      "loss": 1.4365,
+      "step": 7276
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.549259662628174,
+      "learning_rate": 4.807903780068729e-05,
+      "loss": 1.5724,
+      "step": 7277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8265427350997925,
+      "learning_rate": 4.807877346021676e-05,
+      "loss": 1.2762,
+      "step": 7278
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5832833051681519,
+      "learning_rate": 4.8078509119746235e-05,
+      "loss": 1.3251,
+      "step": 7279
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7787108421325684,
+      "learning_rate": 4.807824477927571e-05,
+      "loss": 1.3474,
+      "step": 7280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8743034601211548,
+      "learning_rate": 4.807798043880518e-05,
+      "loss": 1.2639,
+      "step": 7281
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5887702703475952,
+      "learning_rate": 4.8077716098334656e-05,
+      "loss": 1.4374,
+      "step": 7282
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.690917730331421,
+      "learning_rate": 4.8077451757864137e-05,
+      "loss": 1.5468,
+      "step": 7283
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.884947419166565,
+      "learning_rate": 4.80771874173936e-05,
+      "loss": 1.5188,
+      "step": 7284
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5554083585739136,
+      "learning_rate": 4.8076923076923084e-05,
+      "loss": 1.368,
+      "step": 7285
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.945695161819458,
+      "learning_rate": 4.807665873645255e-05,
+      "loss": 1.4277,
+      "step": 7286
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6995351314544678,
+      "learning_rate": 4.8076394395982024e-05,
+      "loss": 1.6111,
+      "step": 7287
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7331725358963013,
+      "learning_rate": 4.8076130055511504e-05,
+      "loss": 1.4132,
+      "step": 7288
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.685289978981018,
+      "learning_rate": 4.807586571504097e-05,
+      "loss": 1.4411,
+      "step": 7289
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4195401668548584,
+      "learning_rate": 4.807560137457045e-05,
+      "loss": 1.1293,
+      "step": 7290
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6508073806762695,
+      "learning_rate": 4.807533703409992e-05,
+      "loss": 1.2717,
+      "step": 7291
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.371819257736206,
+      "learning_rate": 4.80750726936294e-05,
+      "loss": 1.2919,
+      "step": 7292
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4758225679397583,
+      "learning_rate": 4.807480835315887e-05,
+      "loss": 1.2998,
+      "step": 7293
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8697779178619385,
+      "learning_rate": 4.8074544012688346e-05,
+      "loss": 1.3143,
+      "step": 7294
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0728275775909424,
+      "learning_rate": 4.807427967221782e-05,
+      "loss": 1.2572,
+      "step": 7295
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5551948547363281,
+      "learning_rate": 4.807401533174729e-05,
+      "loss": 1.2298,
+      "step": 7296
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7828996181488037,
+      "learning_rate": 4.8073750991276767e-05,
+      "loss": 1.35,
+      "step": 7297
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.900159239768982,
+      "learning_rate": 4.807348665080624e-05,
+      "loss": 1.3915,
+      "step": 7298
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.631771445274353,
+      "learning_rate": 4.8073222310335714e-05,
+      "loss": 1.3252,
+      "step": 7299
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8848572969436646,
+      "learning_rate": 4.807295796986519e-05,
+      "loss": 1.5439,
+      "step": 7300
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.552696704864502,
+      "learning_rate": 4.807269362939466e-05,
+      "loss": 1.4215,
+      "step": 7301
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0572080612182617,
+      "learning_rate": 4.8072429288924134e-05,
+      "loss": 1.408,
+      "step": 7302
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.846600890159607,
+      "learning_rate": 4.8072164948453615e-05,
+      "loss": 1.353,
+      "step": 7303
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9764232635498047,
+      "learning_rate": 4.807190060798308e-05,
+      "loss": 1.5651,
+      "step": 7304
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.70538330078125,
+      "learning_rate": 4.807163626751256e-05,
+      "loss": 1.5355,
+      "step": 7305
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7254866361618042,
+      "learning_rate": 4.807137192704203e-05,
+      "loss": 1.3978,
+      "step": 7306
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5385953187942505,
+      "learning_rate": 4.807110758657151e-05,
+      "loss": 1.4756,
+      "step": 7307
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1012251377105713,
+      "learning_rate": 4.807084324610098e-05,
+      "loss": 1.3906,
+      "step": 7308
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.34948992729187,
+      "learning_rate": 4.8070578905630456e-05,
+      "loss": 1.5577,
+      "step": 7309
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4827839136123657,
+      "learning_rate": 4.807031456515993e-05,
+      "loss": 1.4059,
+      "step": 7310
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2063448429107666,
+      "learning_rate": 4.80700502246894e-05,
+      "loss": 1.4088,
+      "step": 7311
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.044931650161743,
+      "learning_rate": 4.806978588421888e-05,
+      "loss": 1.3284,
+      "step": 7312
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6885443925857544,
+      "learning_rate": 4.806952154374835e-05,
+      "loss": 1.4609,
+      "step": 7313
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.385583519935608,
+      "learning_rate": 4.8069257203277824e-05,
+      "loss": 1.2219,
+      "step": 7314
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.975794792175293,
+      "learning_rate": 4.80689928628073e-05,
+      "loss": 1.5184,
+      "step": 7315
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9418368339538574,
+      "learning_rate": 4.806872852233677e-05,
+      "loss": 1.4946,
+      "step": 7316
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.281878709793091,
+      "learning_rate": 4.8068464181866245e-05,
+      "loss": 1.4351,
+      "step": 7317
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9797919988632202,
+      "learning_rate": 4.8068199841395725e-05,
+      "loss": 1.6059,
+      "step": 7318
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8876811265945435,
+      "learning_rate": 4.806793550092519e-05,
+      "loss": 1.6852,
+      "step": 7319
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7518223524093628,
+      "learning_rate": 4.806767116045467e-05,
+      "loss": 1.46,
+      "step": 7320
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5519944429397583,
+      "learning_rate": 4.806740681998414e-05,
+      "loss": 1.4556,
+      "step": 7321
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4420130252838135,
+      "learning_rate": 4.806714247951362e-05,
+      "loss": 1.2943,
+      "step": 7322
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.539935827255249,
+      "learning_rate": 4.806687813904309e-05,
+      "loss": 1.3909,
+      "step": 7323
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0857157707214355,
+      "learning_rate": 4.806661379857256e-05,
+      "loss": 1.5248,
+      "step": 7324
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6335643529891968,
+      "learning_rate": 4.806634945810204e-05,
+      "loss": 1.3102,
+      "step": 7325
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4977222681045532,
+      "learning_rate": 4.806608511763151e-05,
+      "loss": 1.3683,
+      "step": 7326
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3530731201171875,
+      "learning_rate": 4.806582077716099e-05,
+      "loss": 1.2706,
+      "step": 7327
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7092397212982178,
+      "learning_rate": 4.806555643669046e-05,
+      "loss": 1.4854,
+      "step": 7328
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7350901365280151,
+      "learning_rate": 4.8065292096219934e-05,
+      "loss": 1.4622,
+      "step": 7329
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8712348937988281,
+      "learning_rate": 4.806502775574941e-05,
+      "loss": 1.1586,
+      "step": 7330
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.539783239364624,
+      "learning_rate": 4.806476341527888e-05,
+      "loss": 1.4987,
+      "step": 7331
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6891108751296997,
+      "learning_rate": 4.8064499074808355e-05,
+      "loss": 1.4447,
+      "step": 7332
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9033584594726562,
+      "learning_rate": 4.806423473433783e-05,
+      "loss": 1.3937,
+      "step": 7333
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7034847736358643,
+      "learning_rate": 4.80639703938673e-05,
+      "loss": 1.4191,
+      "step": 7334
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4024947881698608,
+      "learning_rate": 4.8063706053396776e-05,
+      "loss": 1.2926,
+      "step": 7335
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3689146041870117,
+      "learning_rate": 4.806344171292625e-05,
+      "loss": 1.424,
+      "step": 7336
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4197765588760376,
+      "learning_rate": 4.806317737245572e-05,
+      "loss": 1.3707,
+      "step": 7337
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3143888711929321,
+      "learning_rate": 4.80629130319852e-05,
+      "loss": 1.2069,
+      "step": 7338
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.848860502243042,
+      "learning_rate": 4.806264869151467e-05,
+      "loss": 1.3463,
+      "step": 7339
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7952373027801514,
+      "learning_rate": 4.806238435104415e-05,
+      "loss": 1.2336,
+      "step": 7340
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1806633472442627,
+      "learning_rate": 4.806212001057362e-05,
+      "loss": 1.5613,
+      "step": 7341
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.77790367603302,
+      "learning_rate": 4.80618556701031e-05,
+      "loss": 1.4221,
+      "step": 7342
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.071497678756714,
+      "learning_rate": 4.806159132963257e-05,
+      "loss": 1.4886,
+      "step": 7343
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2625863552093506,
+      "learning_rate": 4.8061326989162045e-05,
+      "loss": 1.1365,
+      "step": 7344
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6681668758392334,
+      "learning_rate": 4.806106264869152e-05,
+      "loss": 1.4374,
+      "step": 7345
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6544904708862305,
+      "learning_rate": 4.806079830822099e-05,
+      "loss": 1.4348,
+      "step": 7346
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.73371422290802,
+      "learning_rate": 4.8060533967750465e-05,
+      "loss": 1.5096,
+      "step": 7347
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.92127525806427,
+      "learning_rate": 4.806026962727994e-05,
+      "loss": 1.5371,
+      "step": 7348
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9075062274932861,
+      "learning_rate": 4.806000528680941e-05,
+      "loss": 1.4674,
+      "step": 7349
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7917553186416626,
+      "learning_rate": 4.8059740946338886e-05,
+      "loss": 1.357,
+      "step": 7350
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1925652027130127,
+      "learning_rate": 4.805947660586836e-05,
+      "loss": 1.4494,
+      "step": 7351
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.184513092041016,
+      "learning_rate": 4.805921226539783e-05,
+      "loss": 1.9968,
+      "step": 7352
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.197944164276123,
+      "learning_rate": 4.8058947924927314e-05,
+      "loss": 1.478,
+      "step": 7353
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5615243911743164,
+      "learning_rate": 4.805868358445678e-05,
+      "loss": 1.3691,
+      "step": 7354
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2547571659088135,
+      "learning_rate": 4.805841924398626e-05,
+      "loss": 1.3124,
+      "step": 7355
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.853527307510376,
+      "learning_rate": 4.805815490351573e-05,
+      "loss": 1.5366,
+      "step": 7356
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5195547342300415,
+      "learning_rate": 4.805789056304521e-05,
+      "loss": 1.2936,
+      "step": 7357
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7326598167419434,
+      "learning_rate": 4.805762622257468e-05,
+      "loss": 1.3019,
+      "step": 7358
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8970108032226562,
+      "learning_rate": 4.805736188210415e-05,
+      "loss": 1.3355,
+      "step": 7359
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4218974113464355,
+      "learning_rate": 4.805709754163363e-05,
+      "loss": 1.7158,
+      "step": 7360
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9940234422683716,
+      "learning_rate": 4.8056833201163095e-05,
+      "loss": 1.5317,
+      "step": 7361
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8425525426864624,
+      "learning_rate": 4.8056568860692576e-05,
+      "loss": 1.5274,
+      "step": 7362
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.029553174972534,
+      "learning_rate": 4.805630452022205e-05,
+      "loss": 1.546,
+      "step": 7363
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.00274658203125,
+      "learning_rate": 4.805604017975152e-05,
+      "loss": 1.394,
+      "step": 7364
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5672940015792847,
+      "learning_rate": 4.8055775839280996e-05,
+      "loss": 1.3651,
+      "step": 7365
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9385260343551636,
+      "learning_rate": 4.805551149881047e-05,
+      "loss": 1.6009,
+      "step": 7366
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4309985637664795,
+      "learning_rate": 4.8055247158339944e-05,
+      "loss": 1.4402,
+      "step": 7367
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3831100463867188,
+      "learning_rate": 4.8054982817869424e-05,
+      "loss": 1.6114,
+      "step": 7368
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0133187770843506,
+      "learning_rate": 4.805471847739889e-05,
+      "loss": 1.5409,
+      "step": 7369
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.302422285079956,
+      "learning_rate": 4.8054454136928364e-05,
+      "loss": 1.4585,
+      "step": 7370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.975061297416687,
+      "learning_rate": 4.805418979645784e-05,
+      "loss": 1.416,
+      "step": 7371
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.734453797340393,
+      "learning_rate": 4.805392545598731e-05,
+      "loss": 1.3336,
+      "step": 7372
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.921558141708374,
+      "learning_rate": 4.805366111551679e-05,
+      "loss": 1.4464,
+      "step": 7373
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.172607898712158,
+      "learning_rate": 4.805339677504626e-05,
+      "loss": 1.376,
+      "step": 7374
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.198543071746826,
+      "learning_rate": 4.805313243457574e-05,
+      "loss": 1.4176,
+      "step": 7375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9933907985687256,
+      "learning_rate": 4.8052868094105206e-05,
+      "loss": 1.3344,
+      "step": 7376
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.53383469581604,
+      "learning_rate": 4.8052603753634686e-05,
+      "loss": 1.3654,
+      "step": 7377
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.332566499710083,
+      "learning_rate": 4.805233941316416e-05,
+      "loss": 1.2336,
+      "step": 7378
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.87027907371521,
+      "learning_rate": 4.805207507269363e-05,
+      "loss": 1.4354,
+      "step": 7379
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4681775569915771,
+      "learning_rate": 4.805181073222311e-05,
+      "loss": 1.3762,
+      "step": 7380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4836305379867554,
+      "learning_rate": 4.805154639175258e-05,
+      "loss": 1.3285,
+      "step": 7381
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6713825464248657,
+      "learning_rate": 4.8051282051282054e-05,
+      "loss": 1.558,
+      "step": 7382
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.825882911682129,
+      "learning_rate": 4.805101771081153e-05,
+      "loss": 1.4672,
+      "step": 7383
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7801287174224854,
+      "learning_rate": 4.8050753370341e-05,
+      "loss": 1.4463,
+      "step": 7384
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.88702917098999,
+      "learning_rate": 4.8050489029870475e-05,
+      "loss": 1.6413,
+      "step": 7385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.983485221862793,
+      "learning_rate": 4.805022468939995e-05,
+      "loss": 1.2839,
+      "step": 7386
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.896464467048645,
+      "learning_rate": 4.804996034892942e-05,
+      "loss": 2.0871,
+      "step": 7387
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9741036891937256,
+      "learning_rate": 4.80496960084589e-05,
+      "loss": 1.3517,
+      "step": 7388
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3123154640197754,
+      "learning_rate": 4.804943166798837e-05,
+      "loss": 1.4391,
+      "step": 7389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6525765657424927,
+      "learning_rate": 4.804916732751785e-05,
+      "loss": 1.323,
+      "step": 7390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.933364987373352,
+      "learning_rate": 4.8048902987047316e-05,
+      "loss": 1.5091,
+      "step": 7391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.631428837776184,
+      "learning_rate": 4.8048638646576796e-05,
+      "loss": 1.3613,
+      "step": 7392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0371923446655273,
+      "learning_rate": 4.804837430610627e-05,
+      "loss": 1.5126,
+      "step": 7393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0092382431030273,
+      "learning_rate": 4.8048109965635744e-05,
+      "loss": 1.6065,
+      "step": 7394
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6812547445297241,
+      "learning_rate": 4.804784562516522e-05,
+      "loss": 1.2778,
+      "step": 7395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6871825456619263,
+      "learning_rate": 4.8047581284694684e-05,
+      "loss": 1.2538,
+      "step": 7396
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5789474248886108,
+      "learning_rate": 4.8047316944224164e-05,
+      "loss": 1.402,
+      "step": 7397
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7479296922683716,
+      "learning_rate": 4.804705260375364e-05,
+      "loss": 1.4153,
+      "step": 7398
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4965126514434814,
+      "learning_rate": 4.804678826328311e-05,
+      "loss": 1.6306,
+      "step": 7399
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4232912063598633,
+      "learning_rate": 4.8046523922812585e-05,
+      "loss": 1.244,
+      "step": 7400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.557340145111084,
+      "learning_rate": 4.804625958234206e-05,
+      "loss": 1.3353,
+      "step": 7401
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7588218450546265,
+      "learning_rate": 4.804599524187153e-05,
+      "loss": 1.1794,
+      "step": 7402
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.558750033378601,
+      "learning_rate": 4.804573090140101e-05,
+      "loss": 1.3757,
+      "step": 7403
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.086927890777588,
+      "learning_rate": 4.804546656093048e-05,
+      "loss": 1.6111,
+      "step": 7404
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0300230979919434,
+      "learning_rate": 4.804520222045995e-05,
+      "loss": 1.9596,
+      "step": 7405
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.925351858139038,
+      "learning_rate": 4.8044937879989426e-05,
+      "loss": 1.4636,
+      "step": 7406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.477538824081421,
+      "learning_rate": 4.80446735395189e-05,
+      "loss": 1.2773,
+      "step": 7407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.609445333480835,
+      "learning_rate": 4.804440919904838e-05,
+      "loss": 1.2225,
+      "step": 7408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6230412721633911,
+      "learning_rate": 4.804414485857785e-05,
+      "loss": 1.415,
+      "step": 7409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.977238655090332,
+      "learning_rate": 4.804388051810733e-05,
+      "loss": 1.4467,
+      "step": 7410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.693172812461853,
+      "learning_rate": 4.8043616177636794e-05,
+      "loss": 1.3385,
+      "step": 7411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.577707290649414,
+      "learning_rate": 4.8043351837166275e-05,
+      "loss": 1.3647,
+      "step": 7412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.604644536972046,
+      "learning_rate": 4.804308749669575e-05,
+      "loss": 1.2707,
+      "step": 7413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1357052326202393,
+      "learning_rate": 4.804282315622522e-05,
+      "loss": 1.2371,
+      "step": 7414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.854992389678955,
+      "learning_rate": 4.8042558815754695e-05,
+      "loss": 1.4199,
+      "step": 7415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7727140188217163,
+      "learning_rate": 4.804229447528417e-05,
+      "loss": 1.527,
+      "step": 7416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7020295858383179,
+      "learning_rate": 4.804203013481364e-05,
+      "loss": 1.3481,
+      "step": 7417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7172715663909912,
+      "learning_rate": 4.8041765794343116e-05,
+      "loss": 1.6343,
+      "step": 7418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5492161512374878,
+      "learning_rate": 4.804150145387259e-05,
+      "loss": 1.1765,
+      "step": 7419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.550789713859558,
+      "learning_rate": 4.804123711340206e-05,
+      "loss": 1.377,
+      "step": 7420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8575446605682373,
+      "learning_rate": 4.804097277293154e-05,
+      "loss": 1.4828,
+      "step": 7421
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8084650039672852,
+      "learning_rate": 4.804070843246101e-05,
+      "loss": 1.3081,
+      "step": 7422
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.901208519935608,
+      "learning_rate": 4.804044409199049e-05,
+      "loss": 1.5029,
+      "step": 7423
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.685544013977051,
+      "learning_rate": 4.804017975151996e-05,
+      "loss": 1.3766,
+      "step": 7424
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0926666259765625,
+      "learning_rate": 4.803991541104944e-05,
+      "loss": 1.9336,
+      "step": 7425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5351511240005493,
+      "learning_rate": 4.8039651070578905e-05,
+      "loss": 1.2011,
+      "step": 7426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6568691730499268,
+      "learning_rate": 4.8039386730108385e-05,
+      "loss": 1.3989,
+      "step": 7427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5586135387420654,
+      "learning_rate": 4.803912238963786e-05,
+      "loss": 1.4333,
+      "step": 7428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7156941890716553,
+      "learning_rate": 4.803885804916733e-05,
+      "loss": 1.4494,
+      "step": 7429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.100403070449829,
+      "learning_rate": 4.8038593708696806e-05,
+      "loss": 1.5855,
+      "step": 7430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1658077239990234,
+      "learning_rate": 4.803832936822627e-05,
+      "loss": 1.5859,
+      "step": 7431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5957328081130981,
+      "learning_rate": 4.803806502775575e-05,
+      "loss": 1.2109,
+      "step": 7432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1666078567504883,
+      "learning_rate": 4.8037800687285226e-05,
+      "loss": 1.5039,
+      "step": 7433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8031564950942993,
+      "learning_rate": 4.80375363468147e-05,
+      "loss": 1.3096,
+      "step": 7434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.438870668411255,
+      "learning_rate": 4.8037272006344174e-05,
+      "loss": 1.4457,
+      "step": 7435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5379197597503662,
+      "learning_rate": 4.803700766587365e-05,
+      "loss": 1.376,
+      "step": 7436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.642365574836731,
+      "learning_rate": 4.803674332540312e-05,
+      "loss": 1.3129,
+      "step": 7437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7426648139953613,
+      "learning_rate": 4.80364789849326e-05,
+      "loss": 1.4835,
+      "step": 7438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6209863424301147,
+      "learning_rate": 4.803621464446207e-05,
+      "loss": 1.2122,
+      "step": 7439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0123260021209717,
+      "learning_rate": 4.803595030399155e-05,
+      "loss": 1.5459,
+      "step": 7440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5180186033248901,
+      "learning_rate": 4.8035685963521015e-05,
+      "loss": 1.4818,
+      "step": 7441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1786861419677734,
+      "learning_rate": 4.803542162305049e-05,
+      "loss": 1.4071,
+      "step": 7442
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8186230659484863,
+      "learning_rate": 4.803515728257997e-05,
+      "loss": 1.485,
+      "step": 7443
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6150798797607422,
+      "learning_rate": 4.8034892942109436e-05,
+      "loss": 1.3608,
+      "step": 7444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.433078646659851,
+      "learning_rate": 4.8034628601638916e-05,
+      "loss": 1.2029,
+      "step": 7445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9313353300094604,
+      "learning_rate": 4.803436426116838e-05,
+      "loss": 1.3427,
+      "step": 7446
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6307525634765625,
+      "learning_rate": 4.803409992069786e-05,
+      "loss": 1.3322,
+      "step": 7447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4289510250091553,
+      "learning_rate": 4.803383558022734e-05,
+      "loss": 1.3361,
+      "step": 7448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2729228734970093,
+      "learning_rate": 4.803357123975681e-05,
+      "loss": 1.1035,
+      "step": 7449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3002662658691406,
+      "learning_rate": 4.8033306899286284e-05,
+      "loss": 1.4071,
+      "step": 7450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8397473096847534,
+      "learning_rate": 4.803304255881576e-05,
+      "loss": 1.3689,
+      "step": 7451
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1310524940490723,
+      "learning_rate": 4.803277821834523e-05,
+      "loss": 1.6113,
+      "step": 7452
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9364436864852905,
+      "learning_rate": 4.8032513877874705e-05,
+      "loss": 1.3785,
+      "step": 7453
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.824694037437439,
+      "learning_rate": 4.803224953740418e-05,
+      "loss": 1.5911,
+      "step": 7454
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5871328115463257,
+      "learning_rate": 4.803198519693365e-05,
+      "loss": 1.3433,
+      "step": 7455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8344688415527344,
+      "learning_rate": 4.8031720856463125e-05,
+      "loss": 1.4972,
+      "step": 7456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7328505516052246,
+      "learning_rate": 4.80314565159926e-05,
+      "loss": 1.2375,
+      "step": 7457
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3825576305389404,
+      "learning_rate": 4.803119217552208e-05,
+      "loss": 1.6596,
+      "step": 7458
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.441467523574829,
+      "learning_rate": 4.8030927835051546e-05,
+      "loss": 1.2813,
+      "step": 7459
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7470983266830444,
+      "learning_rate": 4.8030663494581026e-05,
+      "loss": 1.4113,
+      "step": 7460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.519273042678833,
+      "learning_rate": 4.803039915411049e-05,
+      "loss": 1.7415,
+      "step": 7461
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.261181592941284,
+      "learning_rate": 4.8030134813639973e-05,
+      "loss": 1.5175,
+      "step": 7462
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6103876829147339,
+      "learning_rate": 4.802987047316945e-05,
+      "loss": 1.3197,
+      "step": 7463
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3591034412384033,
+      "learning_rate": 4.802960613269892e-05,
+      "loss": 1.2236,
+      "step": 7464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8350086212158203,
+      "learning_rate": 4.8029341792228394e-05,
+      "loss": 1.4509,
+      "step": 7465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1787972450256348,
+      "learning_rate": 4.802907745175787e-05,
+      "loss": 1.5158,
+      "step": 7466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.546413540840149,
+      "learning_rate": 4.802881311128734e-05,
+      "loss": 1.3005,
+      "step": 7467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8614282608032227,
+      "learning_rate": 4.8028548770816815e-05,
+      "loss": 1.3768,
+      "step": 7468
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.671194076538086,
+      "learning_rate": 4.802828443034629e-05,
+      "loss": 1.4081,
+      "step": 7469
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.347003936767578,
+      "learning_rate": 4.802802008987576e-05,
+      "loss": 1.3758,
+      "step": 7470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6933809518814087,
+      "learning_rate": 4.8027755749405236e-05,
+      "loss": 1.4058,
+      "step": 7471
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7203872203826904,
+      "learning_rate": 4.802749140893471e-05,
+      "loss": 1.3324,
+      "step": 7472
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.14042592048645,
+      "learning_rate": 4.802722706846419e-05,
+      "loss": 1.3902,
+      "step": 7473
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3384121656417847,
+      "learning_rate": 4.8026962727993656e-05,
+      "loss": 1.2351,
+      "step": 7474
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.50022029876709,
+      "learning_rate": 4.802669838752314e-05,
+      "loss": 1.7059,
+      "step": 7475
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8446192741394043,
+      "learning_rate": 4.8026434047052603e-05,
+      "loss": 1.4895,
+      "step": 7476
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6634323596954346,
+      "learning_rate": 4.802616970658208e-05,
+      "loss": 1.2616,
+      "step": 7477
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4265505075454712,
+      "learning_rate": 4.802590536611156e-05,
+      "loss": 1.3899,
+      "step": 7478
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.844877004623413,
+      "learning_rate": 4.8025641025641024e-05,
+      "loss": 1.2735,
+      "step": 7479
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.630684494972229,
+      "learning_rate": 4.8025376685170505e-05,
+      "loss": 1.3932,
+      "step": 7480
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9155877828598022,
+      "learning_rate": 4.802511234469997e-05,
+      "loss": 1.3672,
+      "step": 7481
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3998808860778809,
+      "learning_rate": 4.802484800422945e-05,
+      "loss": 1.2436,
+      "step": 7482
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1293375492095947,
+      "learning_rate": 4.8024583663758925e-05,
+      "loss": 1.6023,
+      "step": 7483
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5609501600265503,
+      "learning_rate": 4.80243193232884e-05,
+      "loss": 1.3368,
+      "step": 7484
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.785558819770813,
+      "learning_rate": 4.802405498281787e-05,
+      "loss": 1.3844,
+      "step": 7485
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3915671110153198,
+      "learning_rate": 4.8023790642347346e-05,
+      "loss": 2.3138,
+      "step": 7486
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.78286612033844,
+      "learning_rate": 4.802352630187682e-05,
+      "loss": 1.4303,
+      "step": 7487
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.584670305252075,
+      "learning_rate": 4.802326196140629e-05,
+      "loss": 1.7493,
+      "step": 7488
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6939003467559814,
+      "learning_rate": 4.802299762093577e-05,
+      "loss": 1.3517,
+      "step": 7489
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1180899143218994,
+      "learning_rate": 4.802273328046524e-05,
+      "loss": 1.328,
+      "step": 7490
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7704777717590332,
+      "learning_rate": 4.8022468939994714e-05,
+      "loss": 1.4,
+      "step": 7491
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9235681295394897,
+      "learning_rate": 4.802220459952419e-05,
+      "loss": 1.3423,
+      "step": 7492
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.028578758239746,
+      "learning_rate": 4.802194025905367e-05,
+      "loss": 1.6355,
+      "step": 7493
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7979148626327515,
+      "learning_rate": 4.8021675918583135e-05,
+      "loss": 1.4543,
+      "step": 7494
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7232439517974854,
+      "learning_rate": 4.8021411578112615e-05,
+      "loss": 1.3867,
+      "step": 7495
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.757546067237854,
+      "learning_rate": 4.802114723764208e-05,
+      "loss": 1.4654,
+      "step": 7496
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.131351947784424,
+      "learning_rate": 4.802088289717156e-05,
+      "loss": 1.495,
+      "step": 7497
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7808847427368164,
+      "learning_rate": 4.8020618556701036e-05,
+      "loss": 1.4097,
+      "step": 7498
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.485181212425232,
+      "learning_rate": 4.802035421623051e-05,
+      "loss": 1.3201,
+      "step": 7499
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5884653329849243,
+      "learning_rate": 4.802008987575998e-05,
+      "loss": 1.2202,
+      "step": 7500
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.054253101348877,
+      "learning_rate": 4.8019825535289456e-05,
+      "loss": 1.6021,
+      "step": 7501
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3532179594039917,
+      "learning_rate": 4.801956119481893e-05,
+      "loss": 1.343,
+      "step": 7502
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.096714735031128,
+      "learning_rate": 4.8019296854348403e-05,
+      "loss": 1.5185,
+      "step": 7503
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9005540609359741,
+      "learning_rate": 4.801903251387788e-05,
+      "loss": 1.5273,
+      "step": 7504
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.836534023284912,
+      "learning_rate": 4.801876817340735e-05,
+      "loss": 1.3993,
+      "step": 7505
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.780027151107788,
+      "learning_rate": 4.8018503832936824e-05,
+      "loss": 1.3991,
+      "step": 7506
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4023189544677734,
+      "learning_rate": 4.80182394924663e-05,
+      "loss": 1.3647,
+      "step": 7507
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6544429063796997,
+      "learning_rate": 4.801797515199578e-05,
+      "loss": 1.3205,
+      "step": 7508
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5654951333999634,
+      "learning_rate": 4.8017710811525245e-05,
+      "loss": 1.3694,
+      "step": 7509
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8994299173355103,
+      "learning_rate": 4.8017446471054725e-05,
+      "loss": 1.6484,
+      "step": 7510
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8488167524337769,
+      "learning_rate": 4.801718213058419e-05,
+      "loss": 1.3982,
+      "step": 7511
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.540892243385315,
+      "learning_rate": 4.801691779011367e-05,
+      "loss": 1.4994,
+      "step": 7512
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9487181901931763,
+      "learning_rate": 4.8016653449643146e-05,
+      "loss": 1.4794,
+      "step": 7513
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1995410919189453,
+      "learning_rate": 4.801638910917261e-05,
+      "loss": 1.5548,
+      "step": 7514
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2155354022979736,
+      "learning_rate": 4.801612476870209e-05,
+      "loss": 2.2103,
+      "step": 7515
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.661885142326355,
+      "learning_rate": 4.801586042823156e-05,
+      "loss": 1.3625,
+      "step": 7516
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6643239259719849,
+      "learning_rate": 4.801559608776104e-05,
+      "loss": 1.3619,
+      "step": 7517
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5294958353042603,
+      "learning_rate": 4.8015331747290514e-05,
+      "loss": 1.3419,
+      "step": 7518
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6821194887161255,
+      "learning_rate": 4.801506740681999e-05,
+      "loss": 1.3868,
+      "step": 7519
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8557958602905273,
+      "learning_rate": 4.801480306634946e-05,
+      "loss": 1.4955,
+      "step": 7520
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.77409029006958,
+      "learning_rate": 4.8014538725878934e-05,
+      "loss": 1.2441,
+      "step": 7521
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.859862208366394,
+      "learning_rate": 4.801427438540841e-05,
+      "loss": 1.3818,
+      "step": 7522
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5394684076309204,
+      "learning_rate": 4.801401004493788e-05,
+      "loss": 1.9516,
+      "step": 7523
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6594419479370117,
+      "learning_rate": 4.8013745704467355e-05,
+      "loss": 2.1876,
+      "step": 7524
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4298745393753052,
+      "learning_rate": 4.801348136399683e-05,
+      "loss": 1.3365,
+      "step": 7525
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0137078762054443,
+      "learning_rate": 4.80132170235263e-05,
+      "loss": 1.5857,
+      "step": 7526
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2302303314208984,
+      "learning_rate": 4.8012952683055776e-05,
+      "loss": 1.6518,
+      "step": 7527
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.736842393875122,
+      "learning_rate": 4.8012688342585256e-05,
+      "loss": 1.6624,
+      "step": 7528
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6151784658432007,
+      "learning_rate": 4.801242400211472e-05,
+      "loss": 1.3166,
+      "step": 7529
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7774077653884888,
+      "learning_rate": 4.8012159661644203e-05,
+      "loss": 1.3105,
+      "step": 7530
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8138293027877808,
+      "learning_rate": 4.801189532117367e-05,
+      "loss": 1.4485,
+      "step": 7531
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.533703327178955,
+      "learning_rate": 4.801163098070315e-05,
+      "loss": 1.3371,
+      "step": 7532
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7805720567703247,
+      "learning_rate": 4.8011366640232624e-05,
+      "loss": 1.5244,
+      "step": 7533
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1067512035369873,
+      "learning_rate": 4.80111022997621e-05,
+      "loss": 1.5148,
+      "step": 7534
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5124688148498535,
+      "learning_rate": 4.801083795929157e-05,
+      "loss": 2.3838,
+      "step": 7535
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0765366554260254,
+      "learning_rate": 4.8010573618821045e-05,
+      "loss": 1.5239,
+      "step": 7536
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6260329484939575,
+      "learning_rate": 4.801030927835052e-05,
+      "loss": 1.3811,
+      "step": 7537
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0578088760375977,
+      "learning_rate": 4.801004493787999e-05,
+      "loss": 1.4823,
+      "step": 7538
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0255024433135986,
+      "learning_rate": 4.8009780597409466e-05,
+      "loss": 1.4689,
+      "step": 7539
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6992701292037964,
+      "learning_rate": 4.800951625693894e-05,
+      "loss": 1.3123,
+      "step": 7540
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8626248836517334,
+      "learning_rate": 4.800925191646841e-05,
+      "loss": 1.4076,
+      "step": 7541
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6105501651763916,
+      "learning_rate": 4.8008987575997886e-05,
+      "loss": 1.3085,
+      "step": 7542
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.543149471282959,
+      "learning_rate": 4.8008723235527367e-05,
+      "loss": 2.4716,
+      "step": 7543
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.383571982383728,
+      "learning_rate": 4.8008458895056833e-05,
+      "loss": 1.3852,
+      "step": 7544
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5121421813964844,
+      "learning_rate": 4.8008194554586314e-05,
+      "loss": 1.4036,
+      "step": 7545
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.639838457107544,
+      "learning_rate": 4.800793021411578e-05,
+      "loss": 1.196,
+      "step": 7546
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.459291934967041,
+      "learning_rate": 4.800766587364526e-05,
+      "loss": 1.2233,
+      "step": 7547
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3529698848724365,
+      "learning_rate": 4.8007401533174734e-05,
+      "loss": 1.4046,
+      "step": 7548
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.84817636013031,
+      "learning_rate": 4.80071371927042e-05,
+      "loss": 1.488,
+      "step": 7549
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.69446063041687,
+      "learning_rate": 4.800687285223368e-05,
+      "loss": 1.6951,
+      "step": 7550
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7314682006835938,
+      "learning_rate": 4.800660851176315e-05,
+      "loss": 1.2782,
+      "step": 7551
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2816076278686523,
+      "learning_rate": 4.800634417129263e-05,
+      "loss": 1.2211,
+      "step": 7552
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.506523609161377,
+      "learning_rate": 4.80060798308221e-05,
+      "loss": 1.4619,
+      "step": 7553
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6262485980987549,
+      "learning_rate": 4.8005815490351576e-05,
+      "loss": 1.3991,
+      "step": 7554
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5496810674667358,
+      "learning_rate": 4.800555114988105e-05,
+      "loss": 1.2134,
+      "step": 7555
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7023919820785522,
+      "learning_rate": 4.800528680941052e-05,
+      "loss": 1.2787,
+      "step": 7556
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6214931011199951,
+      "learning_rate": 4.8005022468939997e-05,
+      "loss": 1.2675,
+      "step": 7557
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3521497249603271,
+      "learning_rate": 4.800475812846948e-05,
+      "loss": 1.2871,
+      "step": 7558
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0240368843078613,
+      "learning_rate": 4.8004493787998944e-05,
+      "loss": 1.3835,
+      "step": 7559
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.621175765991211,
+      "learning_rate": 4.800422944752842e-05,
+      "loss": 1.3177,
+      "step": 7560
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6762818098068237,
+      "learning_rate": 4.800396510705789e-05,
+      "loss": 1.421,
+      "step": 7561
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1936347484588623,
+      "learning_rate": 4.8003700766587364e-05,
+      "loss": 1.535,
+      "step": 7562
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.125302791595459,
+      "learning_rate": 4.8003436426116845e-05,
+      "loss": 1.4998,
+      "step": 7563
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.995361566543579,
+      "learning_rate": 4.800317208564631e-05,
+      "loss": 1.5523,
+      "step": 7564
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9417046308517456,
+      "learning_rate": 4.800290774517579e-05,
+      "loss": 1.4697,
+      "step": 7565
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5189800262451172,
+      "learning_rate": 4.800264340470526e-05,
+      "loss": 1.1542,
+      "step": 7566
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2141001224517822,
+      "learning_rate": 4.800237906423474e-05,
+      "loss": 1.5569,
+      "step": 7567
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.277803421020508,
+      "learning_rate": 4.800211472376421e-05,
+      "loss": 1.4298,
+      "step": 7568
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.59273099899292,
+      "learning_rate": 4.8001850383293686e-05,
+      "loss": 1.6217,
+      "step": 7569
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6144715547561646,
+      "learning_rate": 4.800158604282316e-05,
+      "loss": 1.4163,
+      "step": 7570
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.843133568763733,
+      "learning_rate": 4.800132170235263e-05,
+      "loss": 1.3632,
+      "step": 7571
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5436187982559204,
+      "learning_rate": 4.800105736188211e-05,
+      "loss": 1.3749,
+      "step": 7572
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0621538162231445,
+      "learning_rate": 4.800079302141158e-05,
+      "loss": 1.5414,
+      "step": 7573
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7131638526916504,
+      "learning_rate": 4.8000528680941054e-05,
+      "loss": 1.4354,
+      "step": 7574
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6701546907424927,
+      "learning_rate": 4.800026434047053e-05,
+      "loss": 1.3072,
+      "step": 7575
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9113389253616333,
+      "learning_rate": 4.8e-05,
+      "loss": 1.4403,
+      "step": 7576
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3719229698181152,
+      "learning_rate": 4.7999735659529475e-05,
+      "loss": 1.2286,
+      "step": 7577
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6066590547561646,
+      "learning_rate": 4.7999471319058955e-05,
+      "loss": 1.3963,
+      "step": 7578
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6810743808746338,
+      "learning_rate": 4.799920697858842e-05,
+      "loss": 1.4989,
+      "step": 7579
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8784130811691284,
+      "learning_rate": 4.79989426381179e-05,
+      "loss": 1.3869,
+      "step": 7580
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8120214939117432,
+      "learning_rate": 4.799867829764737e-05,
+      "loss": 1.4473,
+      "step": 7581
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7802417278289795,
+      "learning_rate": 4.799841395717685e-05,
+      "loss": 1.4194,
+      "step": 7582
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.500696063041687,
+      "learning_rate": 4.799814961670632e-05,
+      "loss": 1.3883,
+      "step": 7583
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6832365989685059,
+      "learning_rate": 4.7997885276235797e-05,
+      "loss": 1.2911,
+      "step": 7584
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8738536834716797,
+      "learning_rate": 4.799762093576527e-05,
+      "loss": 1.3746,
+      "step": 7585
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.147183656692505,
+      "learning_rate": 4.799735659529474e-05,
+      "loss": 1.3204,
+      "step": 7586
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.647191047668457,
+      "learning_rate": 4.799709225482422e-05,
+      "loss": 1.4804,
+      "step": 7587
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7538909912109375,
+      "learning_rate": 4.799682791435369e-05,
+      "loss": 1.4438,
+      "step": 7588
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5753233432769775,
+      "learning_rate": 4.7996563573883164e-05,
+      "loss": 1.3932,
+      "step": 7589
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7693097591400146,
+      "learning_rate": 4.799629923341264e-05,
+      "loss": 1.3691,
+      "step": 7590
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5687614679336548,
+      "learning_rate": 4.799603489294211e-05,
+      "loss": 1.3328,
+      "step": 7591
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6943190097808838,
+      "learning_rate": 4.7995770552471585e-05,
+      "loss": 1.4742,
+      "step": 7592
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.30804705619812,
+      "learning_rate": 4.7995506212001065e-05,
+      "loss": 1.5157,
+      "step": 7593
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9561771154403687,
+      "learning_rate": 4.799524187153053e-05,
+      "loss": 1.3579,
+      "step": 7594
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6547248363494873,
+      "learning_rate": 4.7994977531060006e-05,
+      "loss": 1.9868,
+      "step": 7595
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2599282264709473,
+      "learning_rate": 4.799471319058948e-05,
+      "loss": 1.3749,
+      "step": 7596
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6154322624206543,
+      "learning_rate": 4.799444885011895e-05,
+      "loss": 1.3452,
+      "step": 7597
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.115452527999878,
+      "learning_rate": 4.799418450964843e-05,
+      "loss": 1.4811,
+      "step": 7598
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.106947183609009,
+      "learning_rate": 4.79939201691779e-05,
+      "loss": 1.3452,
+      "step": 7599
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.478050708770752,
+      "learning_rate": 4.799365582870738e-05,
+      "loss": 1.3356,
+      "step": 7600
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9673182964324951,
+      "learning_rate": 4.799339148823685e-05,
+      "loss": 1.8679,
+      "step": 7601
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7571743726730347,
+      "learning_rate": 4.799312714776633e-05,
+      "loss": 1.4318,
+      "step": 7602
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.717800498008728,
+      "learning_rate": 4.79928628072958e-05,
+      "loss": 1.4147,
+      "step": 7603
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6723581552505493,
+      "learning_rate": 4.7992598466825275e-05,
+      "loss": 1.4022,
+      "step": 7604
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7014636993408203,
+      "learning_rate": 4.799233412635475e-05,
+      "loss": 1.3246,
+      "step": 7605
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8751509189605713,
+      "learning_rate": 4.799206978588422e-05,
+      "loss": 1.4102,
+      "step": 7606
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.365039110183716,
+      "learning_rate": 4.7991805445413695e-05,
+      "loss": 1.5421,
+      "step": 7607
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.276838541030884,
+      "learning_rate": 4.799154110494317e-05,
+      "loss": 1.5825,
+      "step": 7608
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4929274320602417,
+      "learning_rate": 4.799127676447264e-05,
+      "loss": 1.2554,
+      "step": 7609
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5191766023635864,
+      "learning_rate": 4.7991012424002116e-05,
+      "loss": 1.326,
+      "step": 7610
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5629055500030518,
+      "learning_rate": 4.799074808353159e-05,
+      "loss": 1.336,
+      "step": 7611
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8778259754180908,
+      "learning_rate": 4.799048374306106e-05,
+      "loss": 1.3331,
+      "step": 7612
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6681157350540161,
+      "learning_rate": 4.7990219402590544e-05,
+      "loss": 1.4444,
+      "step": 7613
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5668891668319702,
+      "learning_rate": 4.798995506212001e-05,
+      "loss": 2.7943,
+      "step": 7614
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.351508855819702,
+      "learning_rate": 4.798969072164949e-05,
+      "loss": 1.5219,
+      "step": 7615
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8971275091171265,
+      "learning_rate": 4.798942638117896e-05,
+      "loss": 1.4023,
+      "step": 7616
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6411654949188232,
+      "learning_rate": 4.798916204070844e-05,
+      "loss": 1.3661,
+      "step": 7617
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.305147886276245,
+      "learning_rate": 4.798889770023791e-05,
+      "loss": 1.5929,
+      "step": 7618
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0938644409179688,
+      "learning_rate": 4.7988633359767385e-05,
+      "loss": 1.5233,
+      "step": 7619
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6306394338607788,
+      "learning_rate": 4.798836901929686e-05,
+      "loss": 1.1656,
+      "step": 7620
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4805653095245361,
+      "learning_rate": 4.798810467882633e-05,
+      "loss": 1.3124,
+      "step": 7621
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6877034902572632,
+      "learning_rate": 4.7987840338355806e-05,
+      "loss": 1.2899,
+      "step": 7622
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4292653799057007,
+      "learning_rate": 4.798757599788528e-05,
+      "loss": 1.251,
+      "step": 7623
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5555830001831055,
+      "learning_rate": 4.798731165741475e-05,
+      "loss": 1.3372,
+      "step": 7624
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0284273624420166,
+      "learning_rate": 4.7987047316944227e-05,
+      "loss": 1.5093,
+      "step": 7625
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7543892860412598,
+      "learning_rate": 4.79867829764737e-05,
+      "loss": 1.4063,
+      "step": 7626
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0114071369171143,
+      "learning_rate": 4.7986518636003174e-05,
+      "loss": 1.4534,
+      "step": 7627
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6104624271392822,
+      "learning_rate": 4.7986254295532654e-05,
+      "loss": 1.3239,
+      "step": 7628
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.083771228790283,
+      "learning_rate": 4.798598995506212e-05,
+      "loss": 1.3927,
+      "step": 7629
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6131572723388672,
+      "learning_rate": 4.79857256145916e-05,
+      "loss": 1.7706,
+      "step": 7630
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.468509554862976,
+      "learning_rate": 4.798546127412107e-05,
+      "loss": 1.4256,
+      "step": 7631
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8162132501602173,
+      "learning_rate": 4.798519693365054e-05,
+      "loss": 1.6264,
+      "step": 7632
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.289329528808594,
+      "learning_rate": 4.798493259318002e-05,
+      "loss": 2.0087,
+      "step": 7633
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9239715337753296,
+      "learning_rate": 4.798466825270949e-05,
+      "loss": 1.3714,
+      "step": 7634
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8246687650680542,
+      "learning_rate": 4.798440391223897e-05,
+      "loss": 1.3671,
+      "step": 7635
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6762226819992065,
+      "learning_rate": 4.7984139571768436e-05,
+      "loss": 1.3784,
+      "step": 7636
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0441293716430664,
+      "learning_rate": 4.7983875231297916e-05,
+      "loss": 1.4335,
+      "step": 7637
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6218457221984863,
+      "learning_rate": 4.798361089082739e-05,
+      "loss": 1.2083,
+      "step": 7638
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9396644830703735,
+      "learning_rate": 4.798334655035686e-05,
+      "loss": 1.4751,
+      "step": 7639
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4780025482177734,
+      "learning_rate": 4.798308220988634e-05,
+      "loss": 1.7561,
+      "step": 7640
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0638489723205566,
+      "learning_rate": 4.798281786941581e-05,
+      "loss": 1.4892,
+      "step": 7641
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.679099202156067,
+      "learning_rate": 4.7982553528945284e-05,
+      "loss": 1.361,
+      "step": 7642
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6904468536376953,
+      "learning_rate": 4.798228918847476e-05,
+      "loss": 1.3593,
+      "step": 7643
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4264014959335327,
+      "learning_rate": 4.798202484800423e-05,
+      "loss": 1.3777,
+      "step": 7644
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8082772493362427,
+      "learning_rate": 4.7981760507533705e-05,
+      "loss": 1.5171,
+      "step": 7645
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.714287519454956,
+      "learning_rate": 4.798149616706318e-05,
+      "loss": 1.294,
+      "step": 7646
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4904377460479736,
+      "learning_rate": 4.798123182659265e-05,
+      "loss": 1.263,
+      "step": 7647
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.123631238937378,
+      "learning_rate": 4.798096748612213e-05,
+      "loss": 1.6802,
+      "step": 7648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0264153480529785,
+      "learning_rate": 4.79807031456516e-05,
+      "loss": 2.3078,
+      "step": 7649
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8571292161941528,
+      "learning_rate": 4.798043880518108e-05,
+      "loss": 1.4092,
+      "step": 7650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8389402627944946,
+      "learning_rate": 4.7980174464710546e-05,
+      "loss": 1.3725,
+      "step": 7651
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7140498161315918,
+      "learning_rate": 4.7979910124240026e-05,
+      "loss": 1.3077,
+      "step": 7652
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.526916742324829,
+      "learning_rate": 4.79796457837695e-05,
+      "loss": 1.4661,
+      "step": 7653
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6983040571212769,
+      "learning_rate": 4.7979381443298974e-05,
+      "loss": 1.3917,
+      "step": 7654
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9564114809036255,
+      "learning_rate": 4.797911710282845e-05,
+      "loss": 1.3638,
+      "step": 7655
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.138476848602295,
+      "learning_rate": 4.797885276235792e-05,
+      "loss": 1.3769,
+      "step": 7656
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0722877979278564,
+      "learning_rate": 4.7978588421887394e-05,
+      "loss": 1.3945,
+      "step": 7657
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2069625854492188,
+      "learning_rate": 4.797832408141687e-05,
+      "loss": 1.4522,
+      "step": 7658
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3917133808135986,
+      "learning_rate": 4.797805974094634e-05,
+      "loss": 1.3383,
+      "step": 7659
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2416815757751465,
+      "learning_rate": 4.7977795400475815e-05,
+      "loss": 1.5093,
+      "step": 7660
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3370164632797241,
+      "learning_rate": 4.797753106000529e-05,
+      "loss": 1.2273,
+      "step": 7661
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6069070100784302,
+      "learning_rate": 4.797726671953476e-05,
+      "loss": 1.341,
+      "step": 7662
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7974236011505127,
+      "learning_rate": 4.797700237906424e-05,
+      "loss": 1.422,
+      "step": 7663
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2699930667877197,
+      "learning_rate": 4.797673803859371e-05,
+      "loss": 1.3652,
+      "step": 7664
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2890301942825317,
+      "learning_rate": 4.797647369812319e-05,
+      "loss": 1.1598,
+      "step": 7665
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.736612319946289,
+      "learning_rate": 4.7976209357652656e-05,
+      "loss": 1.4844,
+      "step": 7666
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.045140027999878,
+      "learning_rate": 4.797594501718213e-05,
+      "loss": 1.4602,
+      "step": 7667
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.016608953475952,
+      "learning_rate": 4.797568067671161e-05,
+      "loss": 1.3699,
+      "step": 7668
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6962610483169556,
+      "learning_rate": 4.797541633624108e-05,
+      "loss": 1.4331,
+      "step": 7669
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6072611808776855,
+      "learning_rate": 4.797515199577056e-05,
+      "loss": 1.4126,
+      "step": 7670
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6149439811706543,
+      "learning_rate": 4.7974887655300024e-05,
+      "loss": 1.4826,
+      "step": 7671
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8316518068313599,
+      "learning_rate": 4.7974623314829505e-05,
+      "loss": 1.4425,
+      "step": 7672
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6573097705841064,
+      "learning_rate": 4.797435897435898e-05,
+      "loss": 1.3766,
+      "step": 7673
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.595613718032837,
+      "learning_rate": 4.797409463388845e-05,
+      "loss": 1.4453,
+      "step": 7674
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0960052013397217,
+      "learning_rate": 4.7973830293417925e-05,
+      "loss": 1.4965,
+      "step": 7675
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9542522430419922,
+      "learning_rate": 4.79735659529474e-05,
+      "loss": 1.4483,
+      "step": 7676
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0121688842773438,
+      "learning_rate": 4.797330161247687e-05,
+      "loss": 1.5028,
+      "step": 7677
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0648133754730225,
+      "learning_rate": 4.7973037272006346e-05,
+      "loss": 1.3277,
+      "step": 7678
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4610246419906616,
+      "learning_rate": 4.797277293153582e-05,
+      "loss": 1.3157,
+      "step": 7679
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7845652103424072,
+      "learning_rate": 4.797250859106529e-05,
+      "loss": 1.354,
+      "step": 7680
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.170057535171509,
+      "learning_rate": 4.797224425059477e-05,
+      "loss": 1.5015,
+      "step": 7681
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3102478981018066,
+      "learning_rate": 4.797197991012424e-05,
+      "loss": 1.2463,
+      "step": 7682
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4439504146575928,
+      "learning_rate": 4.797171556965372e-05,
+      "loss": 1.5078,
+      "step": 7683
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8985106945037842,
+      "learning_rate": 4.797145122918319e-05,
+      "loss": 1.5345,
+      "step": 7684
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.887160062789917,
+      "learning_rate": 4.797118688871267e-05,
+      "loss": 1.3647,
+      "step": 7685
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.064239978790283,
+      "learning_rate": 4.7970922548242135e-05,
+      "loss": 1.5382,
+      "step": 7686
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5530002117156982,
+      "learning_rate": 4.7970658207771615e-05,
+      "loss": 1.5893,
+      "step": 7687
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6900372505187988,
+      "learning_rate": 4.797039386730109e-05,
+      "loss": 1.461,
+      "step": 7688
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7267276048660278,
+      "learning_rate": 4.797012952683056e-05,
+      "loss": 1.4114,
+      "step": 7689
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0175037384033203,
+      "learning_rate": 4.7969865186360036e-05,
+      "loss": 1.459,
+      "step": 7690
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6127012968063354,
+      "learning_rate": 4.796960084588951e-05,
+      "loss": 1.4027,
+      "step": 7691
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8528046607971191,
+      "learning_rate": 4.796933650541898e-05,
+      "loss": 1.4041,
+      "step": 7692
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8711613416671753,
+      "learning_rate": 4.7969072164948456e-05,
+      "loss": 1.536,
+      "step": 7693
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6387568712234497,
+      "learning_rate": 4.796880782447793e-05,
+      "loss": 1.5612,
+      "step": 7694
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8587135076522827,
+      "learning_rate": 4.7968543484007404e-05,
+      "loss": 1.5154,
+      "step": 7695
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6095515489578247,
+      "learning_rate": 4.796827914353688e-05,
+      "loss": 1.3533,
+      "step": 7696
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6622730493545532,
+      "learning_rate": 4.796801480306635e-05,
+      "loss": 1.4308,
+      "step": 7697
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.480708360671997,
+      "learning_rate": 4.7967750462595824e-05,
+      "loss": 1.4302,
+      "step": 7698
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7593650817871094,
+      "learning_rate": 4.79674861221253e-05,
+      "loss": 1.518,
+      "step": 7699
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2369784116744995,
+      "learning_rate": 4.796722178165478e-05,
+      "loss": 1.2477,
+      "step": 7700
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6227396726608276,
+      "learning_rate": 4.7966957441184245e-05,
+      "loss": 1.4396,
+      "step": 7701
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6857655048370361,
+      "learning_rate": 4.7966693100713725e-05,
+      "loss": 1.3734,
+      "step": 7702
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6575562953948975,
+      "learning_rate": 4.796642876024319e-05,
+      "loss": 1.4042,
+      "step": 7703
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6635247468948364,
+      "learning_rate": 4.7966164419772666e-05,
+      "loss": 1.4342,
+      "step": 7704
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8172332048416138,
+      "learning_rate": 4.7965900079302146e-05,
+      "loss": 1.4696,
+      "step": 7705
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.251922130584717,
+      "learning_rate": 4.796563573883161e-05,
+      "loss": 1.2821,
+      "step": 7706
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5769984722137451,
+      "learning_rate": 4.796537139836109e-05,
+      "loss": 1.3064,
+      "step": 7707
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.587831735610962,
+      "learning_rate": 4.796510705789056e-05,
+      "loss": 1.621,
+      "step": 7708
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5544105768203735,
+      "learning_rate": 4.796484271742004e-05,
+      "loss": 1.4122,
+      "step": 7709
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.732252597808838,
+      "learning_rate": 4.7964578376949514e-05,
+      "loss": 1.5649,
+      "step": 7710
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1846671104431152,
+      "learning_rate": 4.796431403647899e-05,
+      "loss": 1.4567,
+      "step": 7711
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.614542007446289,
+      "learning_rate": 4.796404969600846e-05,
+      "loss": 1.4194,
+      "step": 7712
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.087869167327881,
+      "learning_rate": 4.7963785355537935e-05,
+      "loss": 1.5883,
+      "step": 7713
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.055661916732788,
+      "learning_rate": 4.796352101506741e-05,
+      "loss": 1.349,
+      "step": 7714
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9301726818084717,
+      "learning_rate": 4.796325667459688e-05,
+      "loss": 1.5279,
+      "step": 7715
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4978641271591187,
+      "learning_rate": 4.7962992334126355e-05,
+      "loss": 1.3231,
+      "step": 7716
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.647397756576538,
+      "learning_rate": 4.796272799365583e-05,
+      "loss": 1.4427,
+      "step": 7717
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0082526206970215,
+      "learning_rate": 4.79624636531853e-05,
+      "loss": 1.3662,
+      "step": 7718
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7322534322738647,
+      "learning_rate": 4.7962199312714776e-05,
+      "loss": 1.3913,
+      "step": 7719
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.064016580581665,
+      "learning_rate": 4.7961934972244256e-05,
+      "loss": 1.3126,
+      "step": 7720
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6737157106399536,
+      "learning_rate": 4.796167063177372e-05,
+      "loss": 1.3822,
+      "step": 7721
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7063541412353516,
+      "learning_rate": 4.7961406291303204e-05,
+      "loss": 1.349,
+      "step": 7722
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0966553688049316,
+      "learning_rate": 4.796114195083267e-05,
+      "loss": 1.4102,
+      "step": 7723
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1346328258514404,
+      "learning_rate": 4.796087761036215e-05,
+      "loss": 1.4642,
+      "step": 7724
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.08709454536438,
+      "learning_rate": 4.7960613269891624e-05,
+      "loss": 1.5312,
+      "step": 7725
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4428825378417969,
+      "learning_rate": 4.79603489294211e-05,
+      "loss": 1.3126,
+      "step": 7726
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6798776388168335,
+      "learning_rate": 4.796008458895057e-05,
+      "loss": 1.262,
+      "step": 7727
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.877023458480835,
+      "learning_rate": 4.7959820248480045e-05,
+      "loss": 1.4535,
+      "step": 7728
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8298981189727783,
+      "learning_rate": 4.795955590800952e-05,
+      "loss": 1.428,
+      "step": 7729
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.197957992553711,
+      "learning_rate": 4.795929156753899e-05,
+      "loss": 1.5478,
+      "step": 7730
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.755316972732544,
+      "learning_rate": 4.7959027227068466e-05,
+      "loss": 1.2964,
+      "step": 7731
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6375712156295776,
+      "learning_rate": 4.795876288659794e-05,
+      "loss": 1.3863,
+      "step": 7732
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6668449640274048,
+      "learning_rate": 4.795849854612741e-05,
+      "loss": 1.2793,
+      "step": 7733
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.730920433998108,
+      "learning_rate": 4.7958234205656886e-05,
+      "loss": 1.4351,
+      "step": 7734
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.213977575302124,
+      "learning_rate": 4.795796986518637e-05,
+      "loss": 1.4438,
+      "step": 7735
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3655446767807007,
+      "learning_rate": 4.7957705524715834e-05,
+      "loss": 1.2237,
+      "step": 7736
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8626534938812256,
+      "learning_rate": 4.7957441184245314e-05,
+      "loss": 1.3522,
+      "step": 7737
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8523716926574707,
+      "learning_rate": 4.795717684377478e-05,
+      "loss": 1.4168,
+      "step": 7738
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.913169503211975,
+      "learning_rate": 4.795691250330426e-05,
+      "loss": 1.5219,
+      "step": 7739
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.794547438621521,
+      "learning_rate": 4.7956648162833735e-05,
+      "loss": 1.3864,
+      "step": 7740
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6026463508605957,
+      "learning_rate": 4.79563838223632e-05,
+      "loss": 1.3981,
+      "step": 7741
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6666468381881714,
+      "learning_rate": 4.795611948189268e-05,
+      "loss": 1.3363,
+      "step": 7742
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7334645986557007,
+      "learning_rate": 4.795585514142215e-05,
+      "loss": 1.3304,
+      "step": 7743
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0270256996154785,
+      "learning_rate": 4.795559080095163e-05,
+      "loss": 1.3185,
+      "step": 7744
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.400888442993164,
+      "learning_rate": 4.79553264604811e-05,
+      "loss": 1.4783,
+      "step": 7745
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.529452919960022,
+      "learning_rate": 4.7955062120010576e-05,
+      "loss": 1.1956,
+      "step": 7746
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.186073064804077,
+      "learning_rate": 4.795479777954005e-05,
+      "loss": 1.4301,
+      "step": 7747
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9989449977874756,
+      "learning_rate": 4.795453343906952e-05,
+      "loss": 1.4445,
+      "step": 7748
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.310826063156128,
+      "learning_rate": 4.7954269098599e-05,
+      "loss": 1.6077,
+      "step": 7749
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.568924903869629,
+      "learning_rate": 4.795400475812847e-05,
+      "loss": 1.4098,
+      "step": 7750
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3460333347320557,
+      "learning_rate": 4.7953740417657944e-05,
+      "loss": 1.3185,
+      "step": 7751
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.054888963699341,
+      "learning_rate": 4.795347607718742e-05,
+      "loss": 1.456,
+      "step": 7752
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5937107801437378,
+      "learning_rate": 4.795321173671689e-05,
+      "loss": 1.3519,
+      "step": 7753
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.001772403717041,
+      "learning_rate": 4.7952947396246365e-05,
+      "loss": 1.5086,
+      "step": 7754
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.353057384490967,
+      "learning_rate": 4.7952683055775845e-05,
+      "loss": 1.5007,
+      "step": 7755
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7383909225463867,
+      "learning_rate": 4.795241871530531e-05,
+      "loss": 1.3626,
+      "step": 7756
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2861275672912598,
+      "learning_rate": 4.795215437483479e-05,
+      "loss": 1.3079,
+      "step": 7757
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2342710494995117,
+      "learning_rate": 4.795189003436426e-05,
+      "loss": 1.365,
+      "step": 7758
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.644469141960144,
+      "learning_rate": 4.795162569389374e-05,
+      "loss": 1.5605,
+      "step": 7759
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7591789960861206,
+      "learning_rate": 4.795136135342321e-05,
+      "loss": 1.3311,
+      "step": 7760
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.618856430053711,
+      "learning_rate": 4.7951097012952686e-05,
+      "loss": 1.4426,
+      "step": 7761
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.855370044708252,
+      "learning_rate": 4.795083267248216e-05,
+      "loss": 1.3985,
+      "step": 7762
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6523032188415527,
+      "learning_rate": 4.7950568332011634e-05,
+      "loss": 1.786,
+      "step": 7763
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2333844900131226,
+      "learning_rate": 4.795030399154111e-05,
+      "loss": 1.2059,
+      "step": 7764
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8383512496948242,
+      "learning_rate": 4.795003965107058e-05,
+      "loss": 1.3986,
+      "step": 7765
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.09069561958313,
+      "learning_rate": 4.7949775310600054e-05,
+      "loss": 1.4574,
+      "step": 7766
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9426777362823486,
+      "learning_rate": 4.794951097012953e-05,
+      "loss": 1.4843,
+      "step": 7767
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4740920066833496,
+      "learning_rate": 4.7949246629659e-05,
+      "loss": 1.5805,
+      "step": 7768
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.363117218017578,
+      "learning_rate": 4.7948982289188475e-05,
+      "loss": 1.6065,
+      "step": 7769
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.749375581741333,
+      "learning_rate": 4.7948717948717955e-05,
+      "loss": 1.3951,
+      "step": 7770
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6813411712646484,
+      "learning_rate": 4.794845360824742e-05,
+      "loss": 1.5592,
+      "step": 7771
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9595232009887695,
+      "learning_rate": 4.79481892677769e-05,
+      "loss": 1.5244,
+      "step": 7772
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.665564775466919,
+      "learning_rate": 4.794792492730637e-05,
+      "loss": 1.3841,
+      "step": 7773
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7099668979644775,
+      "learning_rate": 4.794766058683585e-05,
+      "loss": 1.539,
+      "step": 7774
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3844704627990723,
+      "learning_rate": 4.794739624636532e-05,
+      "loss": 1.3875,
+      "step": 7775
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6250503063201904,
+      "learning_rate": 4.794713190589479e-05,
+      "loss": 1.2768,
+      "step": 7776
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4919569492340088,
+      "learning_rate": 4.794686756542427e-05,
+      "loss": 1.4305,
+      "step": 7777
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4643789529800415,
+      "learning_rate": 4.794660322495374e-05,
+      "loss": 1.2428,
+      "step": 7778
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6816117763519287,
+      "learning_rate": 4.794633888448322e-05,
+      "loss": 1.499,
+      "step": 7779
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1128201484680176,
+      "learning_rate": 4.794607454401269e-05,
+      "loss": 1.5106,
+      "step": 7780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8911415338516235,
+      "learning_rate": 4.7945810203542165e-05,
+      "loss": 1.5216,
+      "step": 7781
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.444931745529175,
+      "learning_rate": 4.794554586307164e-05,
+      "loss": 2.4058,
+      "step": 7782
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.031721591949463,
+      "learning_rate": 4.794528152260111e-05,
+      "loss": 1.592,
+      "step": 7783
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.689886212348938,
+      "learning_rate": 4.7945017182130585e-05,
+      "loss": 1.4233,
+      "step": 7784
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2977964878082275,
+      "learning_rate": 4.794475284166006e-05,
+      "loss": 1.5573,
+      "step": 7785
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6014795303344727,
+      "learning_rate": 4.794448850118953e-05,
+      "loss": 1.6579,
+      "step": 7786
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4291749000549316,
+      "learning_rate": 4.7944224160719006e-05,
+      "loss": 1.7358,
+      "step": 7787
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4474140405654907,
+      "learning_rate": 4.794395982024848e-05,
+      "loss": 1.3605,
+      "step": 7788
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7285571098327637,
+      "learning_rate": 4.794369547977795e-05,
+      "loss": 1.3989,
+      "step": 7789
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9054203033447266,
+      "learning_rate": 4.7943431139307433e-05,
+      "loss": 1.5566,
+      "step": 7790
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6477717161178589,
+      "learning_rate": 4.79431667988369e-05,
+      "loss": 1.4649,
+      "step": 7791
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7155696153640747,
+      "learning_rate": 4.794290245836638e-05,
+      "loss": 1.3696,
+      "step": 7792
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.462106227874756,
+      "learning_rate": 4.794263811789585e-05,
+      "loss": 1.5243,
+      "step": 7793
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7765305042266846,
+      "learning_rate": 4.794237377742533e-05,
+      "loss": 1.2236,
+      "step": 7794
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.469255805015564,
+      "learning_rate": 4.79421094369548e-05,
+      "loss": 1.3814,
+      "step": 7795
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7083935737609863,
+      "learning_rate": 4.7941845096484275e-05,
+      "loss": 1.3426,
+      "step": 7796
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.373934268951416,
+      "learning_rate": 4.794158075601375e-05,
+      "loss": 1.2549,
+      "step": 7797
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.250418186187744,
+      "learning_rate": 4.794131641554322e-05,
+      "loss": 1.6324,
+      "step": 7798
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.484714388847351,
+      "learning_rate": 4.7941052075072696e-05,
+      "loss": 1.3567,
+      "step": 7799
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7283401489257812,
+      "learning_rate": 4.794078773460217e-05,
+      "loss": 1.5262,
+      "step": 7800
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2799875736236572,
+      "learning_rate": 4.794052339413164e-05,
+      "loss": 2.1497,
+      "step": 7801
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4556074142456055,
+      "learning_rate": 4.7940259053661116e-05,
+      "loss": 1.3044,
+      "step": 7802
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6997997760772705,
+      "learning_rate": 4.793999471319059e-05,
+      "loss": 1.4491,
+      "step": 7803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.051269292831421,
+      "learning_rate": 4.7939730372720063e-05,
+      "loss": 1.45,
+      "step": 7804
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6978527307510376,
+      "learning_rate": 4.7939466032249544e-05,
+      "loss": 1.2753,
+      "step": 7805
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5851155519485474,
+      "learning_rate": 4.793920169177901e-05,
+      "loss": 1.3795,
+      "step": 7806
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7675386667251587,
+      "learning_rate": 4.793893735130849e-05,
+      "loss": 1.5551,
+      "step": 7807
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.171959400177002,
+      "learning_rate": 4.793867301083796e-05,
+      "loss": 1.3917,
+      "step": 7808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.813775897026062,
+      "learning_rate": 4.793840867036744e-05,
+      "loss": 1.3682,
+      "step": 7809
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.661798119544983,
+      "learning_rate": 4.793814432989691e-05,
+      "loss": 1.4967,
+      "step": 7810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.722652792930603,
+      "learning_rate": 4.7937879989426385e-05,
+      "loss": 1.3913,
+      "step": 7811
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.530395269393921,
+      "learning_rate": 4.793761564895586e-05,
+      "loss": 1.5934,
+      "step": 7812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7117249965667725,
+      "learning_rate": 4.7937351308485326e-05,
+      "loss": 1.4297,
+      "step": 7813
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8461365699768066,
+      "learning_rate": 4.7937086968014806e-05,
+      "loss": 1.341,
+      "step": 7814
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6335623264312744,
+      "learning_rate": 4.793682262754428e-05,
+      "loss": 1.2759,
+      "step": 7815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.761736512184143,
+      "learning_rate": 4.793655828707375e-05,
+      "loss": 1.4419,
+      "step": 7816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5849696397781372,
+      "learning_rate": 4.793629394660323e-05,
+      "loss": 1.3874,
+      "step": 7817
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7939635515213013,
+      "learning_rate": 4.79360296061327e-05,
+      "loss": 1.4318,
+      "step": 7818
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7996982336044312,
+      "learning_rate": 4.7935765265662174e-05,
+      "loss": 1.4684,
+      "step": 7819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.52805757522583,
+      "learning_rate": 4.7935500925191654e-05,
+      "loss": 1.3515,
+      "step": 7820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4186776876449585,
+      "learning_rate": 4.793523658472112e-05,
+      "loss": 1.3415,
+      "step": 7821
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.615703821182251,
+      "learning_rate": 4.7934972244250595e-05,
+      "loss": 1.3683,
+      "step": 7822
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6783297061920166,
+      "learning_rate": 4.793470790378007e-05,
+      "loss": 1.3432,
+      "step": 7823
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2005667686462402,
+      "learning_rate": 4.793444356330954e-05,
+      "loss": 1.3804,
+      "step": 7824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4147332906723022,
+      "learning_rate": 4.793417922283902e-05,
+      "loss": 1.3645,
+      "step": 7825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1372897624969482,
+      "learning_rate": 4.793391488236849e-05,
+      "loss": 1.6494,
+      "step": 7826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6363667249679565,
+      "learning_rate": 4.793365054189797e-05,
+      "loss": 1.3771,
+      "step": 7827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7421125173568726,
+      "learning_rate": 4.7933386201427436e-05,
+      "loss": 1.5224,
+      "step": 7828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8436076641082764,
+      "learning_rate": 4.7933121860956916e-05,
+      "loss": 1.4442,
+      "step": 7829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.878571629524231,
+      "learning_rate": 4.793285752048639e-05,
+      "loss": 1.5828,
+      "step": 7830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6789835691452026,
+      "learning_rate": 4.7932593180015863e-05,
+      "loss": 1.4264,
+      "step": 7831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.016465425491333,
+      "learning_rate": 4.793232883954534e-05,
+      "loss": 1.6478,
+      "step": 7832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5049749612808228,
+      "learning_rate": 4.793206449907481e-05,
+      "loss": 1.2255,
+      "step": 7833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2175137996673584,
+      "learning_rate": 4.7931800158604284e-05,
+      "loss": 1.4727,
+      "step": 7834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6115080118179321,
+      "learning_rate": 4.793153581813376e-05,
+      "loss": 1.3823,
+      "step": 7835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.041806936264038,
+      "learning_rate": 4.793127147766323e-05,
+      "loss": 1.4141,
+      "step": 7836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6653575897216797,
+      "learning_rate": 4.7931007137192705e-05,
+      "loss": 1.3089,
+      "step": 7837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7942463159561157,
+      "learning_rate": 4.793074279672218e-05,
+      "loss": 1.3569,
+      "step": 7838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5390403270721436,
+      "learning_rate": 4.793047845625165e-05,
+      "loss": 1.3185,
+      "step": 7839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8173158168792725,
+      "learning_rate": 4.793021411578113e-05,
+      "loss": 1.2873,
+      "step": 7840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.52585768699646,
+      "learning_rate": 4.79299497753106e-05,
+      "loss": 2.0069,
+      "step": 7841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.116159439086914,
+      "learning_rate": 4.792968543484008e-05,
+      "loss": 1.4805,
+      "step": 7842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.233954429626465,
+      "learning_rate": 4.7929421094369546e-05,
+      "loss": 1.5357,
+      "step": 7843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.908521056175232,
+      "learning_rate": 4.792915675389903e-05,
+      "loss": 1.4777,
+      "step": 7844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6433674097061157,
+      "learning_rate": 4.79288924134285e-05,
+      "loss": 1.2782,
+      "step": 7845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5719077587127686,
+      "learning_rate": 4.7928628072957974e-05,
+      "loss": 2.2388,
+      "step": 7846
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0320262908935547,
+      "learning_rate": 4.792836373248745e-05,
+      "loss": 1.4588,
+      "step": 7847
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9671610593795776,
+      "learning_rate": 4.7928099392016914e-05,
+      "loss": 1.5007,
+      "step": 7848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4089150428771973,
+      "learning_rate": 4.7927835051546394e-05,
+      "loss": 1.2939,
+      "step": 7849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1510958671569824,
+      "learning_rate": 4.792757071107587e-05,
+      "loss": 1.5367,
+      "step": 7850
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.003730297088623,
+      "learning_rate": 4.792730637060534e-05,
+      "loss": 1.409,
+      "step": 7851
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.249903678894043,
+      "learning_rate": 4.7927042030134815e-05,
+      "loss": 1.4069,
+      "step": 7852
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2254221439361572,
+      "learning_rate": 4.792677768966429e-05,
+      "loss": 2.428,
+      "step": 7853
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5402023792266846,
+      "learning_rate": 4.792651334919376e-05,
+      "loss": 1.3943,
+      "step": 7854
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4857054948806763,
+      "learning_rate": 4.792624900872324e-05,
+      "loss": 1.4199,
+      "step": 7855
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8432390689849854,
+      "learning_rate": 4.792598466825271e-05,
+      "loss": 1.4928,
+      "step": 7856
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7518781423568726,
+      "learning_rate": 4.792572032778219e-05,
+      "loss": 1.4723,
+      "step": 7857
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8388631343841553,
+      "learning_rate": 4.792545598731166e-05,
+      "loss": 1.3813,
+      "step": 7858
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.820644736289978,
+      "learning_rate": 4.792519164684113e-05,
+      "loss": 1.3891,
+      "step": 7859
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0197649002075195,
+      "learning_rate": 4.792492730637061e-05,
+      "loss": 1.3844,
+      "step": 7860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6017169952392578,
+      "learning_rate": 4.792466296590008e-05,
+      "loss": 1.4774,
+      "step": 7861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6232824325561523,
+      "learning_rate": 4.792439862542956e-05,
+      "loss": 1.3243,
+      "step": 7862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7220709323883057,
+      "learning_rate": 4.7924134284959024e-05,
+      "loss": 1.2995,
+      "step": 7863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7662776708602905,
+      "learning_rate": 4.7923869944488505e-05,
+      "loss": 1.5275,
+      "step": 7864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5594029426574707,
+      "learning_rate": 4.792360560401798e-05,
+      "loss": 1.3068,
+      "step": 7865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5694105625152588,
+      "learning_rate": 4.792334126354745e-05,
+      "loss": 1.3493,
+      "step": 7866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4185173511505127,
+      "learning_rate": 4.7923076923076926e-05,
+      "loss": 1.2749,
+      "step": 7867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8340601921081543,
+      "learning_rate": 4.79228125826064e-05,
+      "loss": 1.391,
+      "step": 7868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7911303043365479,
+      "learning_rate": 4.792254824213587e-05,
+      "loss": 1.5402,
+      "step": 7869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9333703517913818,
+      "learning_rate": 4.7922283901665346e-05,
+      "loss": 1.3741,
+      "step": 7870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9112688302993774,
+      "learning_rate": 4.792201956119482e-05,
+      "loss": 1.3911,
+      "step": 7871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.555514097213745,
+      "learning_rate": 4.7921755220724293e-05,
+      "loss": 1.6129,
+      "step": 7872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6918325424194336,
+      "learning_rate": 4.792149088025377e-05,
+      "loss": 1.4077,
+      "step": 7873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5825891494750977,
+      "learning_rate": 4.792122653978324e-05,
+      "loss": 1.3595,
+      "step": 7874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8335739374160767,
+      "learning_rate": 4.792096219931272e-05,
+      "loss": 1.3817,
+      "step": 7875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5327516794204712,
+      "learning_rate": 4.792069785884219e-05,
+      "loss": 1.365,
+      "step": 7876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9348586797714233,
+      "learning_rate": 4.792043351837167e-05,
+      "loss": 1.3758,
+      "step": 7877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3617485761642456,
+      "learning_rate": 4.7920169177901135e-05,
+      "loss": 1.3649,
+      "step": 7878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7141358852386475,
+      "learning_rate": 4.7919904837430615e-05,
+      "loss": 1.3491,
+      "step": 7879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5741958618164062,
+      "learning_rate": 4.791964049696009e-05,
+      "loss": 2.2931,
+      "step": 7880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6785244941711426,
+      "learning_rate": 4.791937615648956e-05,
+      "loss": 1.4353,
+      "step": 7881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5229851007461548,
+      "learning_rate": 4.7919111816019036e-05,
+      "loss": 1.3662,
+      "step": 7882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9916540384292603,
+      "learning_rate": 4.791884747554851e-05,
+      "loss": 1.2865,
+      "step": 7883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.528250217437744,
+      "learning_rate": 4.791858313507798e-05,
+      "loss": 1.5779,
+      "step": 7884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4746885299682617,
+      "learning_rate": 4.7918318794607457e-05,
+      "loss": 1.4774,
+      "step": 7885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9873285293579102,
+      "learning_rate": 4.791805445413693e-05,
+      "loss": 1.3807,
+      "step": 7886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0445945262908936,
+      "learning_rate": 4.7917790113666404e-05,
+      "loss": 1.4417,
+      "step": 7887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7752615213394165,
+      "learning_rate": 4.791752577319588e-05,
+      "loss": 1.3249,
+      "step": 7888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6283328533172607,
+      "learning_rate": 4.791726143272535e-05,
+      "loss": 1.3561,
+      "step": 7889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.64657461643219,
+      "learning_rate": 4.791699709225483e-05,
+      "loss": 1.4867,
+      "step": 7890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6178034543991089,
+      "learning_rate": 4.79167327517843e-05,
+      "loss": 1.3881,
+      "step": 7891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6223396062850952,
+      "learning_rate": 4.791646841131378e-05,
+      "loss": 1.3988,
+      "step": 7892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8224855661392212,
+      "learning_rate": 4.7916204070843245e-05,
+      "loss": 1.4099,
+      "step": 7893
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0050926208496094,
+      "learning_rate": 4.791593973037272e-05,
+      "loss": 2.6359,
+      "step": 7894
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.378525733947754,
+      "learning_rate": 4.79156753899022e-05,
+      "loss": 1.1461,
+      "step": 7895
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6890087127685547,
+      "learning_rate": 4.7915411049431666e-05,
+      "loss": 1.314,
+      "step": 7896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.603840708732605,
+      "learning_rate": 4.7915146708961146e-05,
+      "loss": 1.331,
+      "step": 7897
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9266732931137085,
+      "learning_rate": 4.791488236849061e-05,
+      "loss": 1.4557,
+      "step": 7898
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8215622901916504,
+      "learning_rate": 4.791461802802009e-05,
+      "loss": 1.3652,
+      "step": 7899
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.460837721824646,
+      "learning_rate": 4.791435368754957e-05,
+      "loss": 1.2467,
+      "step": 7900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6477514505386353,
+      "learning_rate": 4.791408934707904e-05,
+      "loss": 1.2981,
+      "step": 7901
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8197810649871826,
+      "learning_rate": 4.7913825006608514e-05,
+      "loss": 1.5501,
+      "step": 7902
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.196084499359131,
+      "learning_rate": 4.791356066613799e-05,
+      "loss": 1.485,
+      "step": 7903
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8380810022354126,
+      "learning_rate": 4.791329632566746e-05,
+      "loss": 1.3783,
+      "step": 7904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.315187692642212,
+      "learning_rate": 4.7913031985196935e-05,
+      "loss": 1.3272,
+      "step": 7905
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4320683479309082,
+      "learning_rate": 4.791276764472641e-05,
+      "loss": 1.3018,
+      "step": 7906
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9525318145751953,
+      "learning_rate": 4.791250330425588e-05,
+      "loss": 1.4223,
+      "step": 7907
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.040900707244873,
+      "learning_rate": 4.7912238963785356e-05,
+      "loss": 1.4732,
+      "step": 7908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6550654172897339,
+      "learning_rate": 4.791197462331483e-05,
+      "loss": 1.4658,
+      "step": 7909
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3974709510803223,
+      "learning_rate": 4.791171028284431e-05,
+      "loss": 1.34,
+      "step": 7910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.580579161643982,
+      "learning_rate": 4.7911445942373776e-05,
+      "loss": 1.4085,
+      "step": 7911
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.957444667816162,
+      "learning_rate": 4.7911181601903257e-05,
+      "loss": 1.2934,
+      "step": 7912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6807941198349,
+      "learning_rate": 4.791091726143272e-05,
+      "loss": 1.337,
+      "step": 7913
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5197594165802002,
+      "learning_rate": 4.7910652920962204e-05,
+      "loss": 1.3674,
+      "step": 7914
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.584311842918396,
+      "learning_rate": 4.791038858049168e-05,
+      "loss": 1.4229,
+      "step": 7915
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8339662551879883,
+      "learning_rate": 4.791012424002115e-05,
+      "loss": 1.4895,
+      "step": 7916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4321317672729492,
+      "learning_rate": 4.7909859899550624e-05,
+      "loss": 1.3585,
+      "step": 7917
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6281780004501343,
+      "learning_rate": 4.79095955590801e-05,
+      "loss": 1.4387,
+      "step": 7918
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5034055709838867,
+      "learning_rate": 4.790933121860957e-05,
+      "loss": 1.7221,
+      "step": 7919
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8324466943740845,
+      "learning_rate": 4.7909066878139045e-05,
+      "loss": 1.4626,
+      "step": 7920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5424115657806396,
+      "learning_rate": 4.790880253766852e-05,
+      "loss": 1.2931,
+      "step": 7921
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.055544376373291,
+      "learning_rate": 4.790853819719799e-05,
+      "loss": 1.3916,
+      "step": 7922
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4082558155059814,
+      "learning_rate": 4.7908273856727466e-05,
+      "loss": 1.3247,
+      "step": 7923
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5213347673416138,
+      "learning_rate": 4.790800951625694e-05,
+      "loss": 1.2878,
+      "step": 7924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.543341040611267,
+      "learning_rate": 4.790774517578642e-05,
+      "loss": 1.2647,
+      "step": 7925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.49587082862854,
+      "learning_rate": 4.7907480835315887e-05,
+      "loss": 1.3119,
+      "step": 7926
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.380164861679077,
+      "learning_rate": 4.790721649484537e-05,
+      "loss": 1.6538,
+      "step": 7927
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0560977458953857,
+      "learning_rate": 4.7906952154374834e-05,
+      "loss": 1.4525,
+      "step": 7928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.303464889526367,
+      "learning_rate": 4.7906687813904314e-05,
+      "loss": 1.6558,
+      "step": 7929
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.656042218208313,
+      "learning_rate": 4.790642347343379e-05,
+      "loss": 1.2323,
+      "step": 7930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5798418521881104,
+      "learning_rate": 4.7906159132963254e-05,
+      "loss": 1.2696,
+      "step": 7931
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.451194167137146,
+      "learning_rate": 4.7905894792492735e-05,
+      "loss": 1.3131,
+      "step": 7932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.462851643562317,
+      "learning_rate": 4.79056304520222e-05,
+      "loss": 1.3505,
+      "step": 7933
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1646435260772705,
+      "learning_rate": 4.790536611155168e-05,
+      "loss": 1.3655,
+      "step": 7934
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4100160598754883,
+      "learning_rate": 4.7905101771081155e-05,
+      "loss": 1.4333,
+      "step": 7935
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8242614269256592,
+      "learning_rate": 4.790483743061063e-05,
+      "loss": 1.3424,
+      "step": 7936
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8949211835861206,
+      "learning_rate": 4.79045730901401e-05,
+      "loss": 1.3713,
+      "step": 7937
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4765461683273315,
+      "learning_rate": 4.7904308749669576e-05,
+      "loss": 1.4109,
+      "step": 7938
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6186432838439941,
+      "learning_rate": 4.790404440919905e-05,
+      "loss": 1.3713,
+      "step": 7939
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.472906231880188,
+      "learning_rate": 4.790378006872852e-05,
+      "loss": 1.3421,
+      "step": 7940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.696186900138855,
+      "learning_rate": 4.7903515728258e-05,
+      "loss": 1.3341,
+      "step": 7941
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6466103792190552,
+      "learning_rate": 4.790325138778747e-05,
+      "loss": 1.2266,
+      "step": 7942
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3000348806381226,
+      "learning_rate": 4.7902987047316944e-05,
+      "loss": 1.1953,
+      "step": 7943
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.830739974975586,
+      "learning_rate": 4.790272270684642e-05,
+      "loss": 1.3518,
+      "step": 7944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9814186096191406,
+      "learning_rate": 4.79024583663759e-05,
+      "loss": 1.3336,
+      "step": 7945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7689309120178223,
+      "learning_rate": 4.7902194025905365e-05,
+      "loss": 1.1879,
+      "step": 7946
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.08500337600708,
+      "learning_rate": 4.7901929685434845e-05,
+      "loss": 1.4154,
+      "step": 7947
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8341050148010254,
+      "learning_rate": 4.790166534496431e-05,
+      "loss": 1.4121,
+      "step": 7948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5473171472549438,
+      "learning_rate": 4.790140100449379e-05,
+      "loss": 1.4044,
+      "step": 7949
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.919596791267395,
+      "learning_rate": 4.7901136664023266e-05,
+      "loss": 1.284,
+      "step": 7950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9843789339065552,
+      "learning_rate": 4.790087232355274e-05,
+      "loss": 1.4601,
+      "step": 7951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.976767897605896,
+      "learning_rate": 4.790060798308221e-05,
+      "loss": 1.4981,
+      "step": 7952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4187376499176025,
+      "learning_rate": 4.7900343642611687e-05,
+      "loss": 1.2194,
+      "step": 7953
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3983583450317383,
+      "learning_rate": 4.790007930214116e-05,
+      "loss": 1.3448,
+      "step": 7954
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6223636865615845,
+      "learning_rate": 4.7899814961670634e-05,
+      "loss": 1.4153,
+      "step": 7955
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6770085096359253,
+      "learning_rate": 4.789955062120011e-05,
+      "loss": 1.4162,
+      "step": 7956
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0790178775787354,
+      "learning_rate": 4.789928628072958e-05,
+      "loss": 1.4606,
+      "step": 7957
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7172802686691284,
+      "learning_rate": 4.7899021940259054e-05,
+      "loss": 1.5547,
+      "step": 7958
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8615210056304932,
+      "learning_rate": 4.789875759978853e-05,
+      "loss": 1.5032,
+      "step": 7959
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.539447546005249,
+      "learning_rate": 4.789849325931801e-05,
+      "loss": 1.4481,
+      "step": 7960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.76738440990448,
+      "learning_rate": 4.7898228918847475e-05,
+      "loss": 1.4134,
+      "step": 7961
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.544714331626892,
+      "learning_rate": 4.7897964578376955e-05,
+      "loss": 1.3002,
+      "step": 7962
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2510929107666016,
+      "learning_rate": 4.789770023790642e-05,
+      "loss": 1.4891,
+      "step": 7963
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0860815048217773,
+      "learning_rate": 4.78974358974359e-05,
+      "loss": 1.4852,
+      "step": 7964
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.032287120819092,
+      "learning_rate": 4.7897171556965376e-05,
+      "loss": 1.5633,
+      "step": 7965
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.580087423324585,
+      "learning_rate": 4.789690721649484e-05,
+      "loss": 1.4163,
+      "step": 7966
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.504023313522339,
+      "learning_rate": 4.789664287602432e-05,
+      "loss": 1.4034,
+      "step": 7967
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8707993030548096,
+      "learning_rate": 4.789637853555379e-05,
+      "loss": 1.3349,
+      "step": 7968
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5471596717834473,
+      "learning_rate": 4.789611419508327e-05,
+      "loss": 1.2335,
+      "step": 7969
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6922895908355713,
+      "learning_rate": 4.7895849854612744e-05,
+      "loss": 1.6217,
+      "step": 7970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4075239896774292,
+      "learning_rate": 4.789558551414222e-05,
+      "loss": 1.3905,
+      "step": 7971
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1879185438156128,
+      "learning_rate": 4.789532117367169e-05,
+      "loss": 1.0904,
+      "step": 7972
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2606509923934937,
+      "learning_rate": 4.7895056833201165e-05,
+      "loss": 1.235,
+      "step": 7973
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.961582660675049,
+      "learning_rate": 4.789479249273064e-05,
+      "loss": 1.5579,
+      "step": 7974
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9594523906707764,
+      "learning_rate": 4.789452815226011e-05,
+      "loss": 1.4551,
+      "step": 7975
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5953409671783447,
+      "learning_rate": 4.7894263811789585e-05,
+      "loss": 1.2927,
+      "step": 7976
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8516865968704224,
+      "learning_rate": 4.789399947131906e-05,
+      "loss": 1.3548,
+      "step": 7977
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8768571615219116,
+      "learning_rate": 4.789373513084853e-05,
+      "loss": 1.5522,
+      "step": 7978
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.344625234603882,
+      "learning_rate": 4.7893470790378006e-05,
+      "loss": 1.4615,
+      "step": 7979
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.818874716758728,
+      "learning_rate": 4.7893206449907486e-05,
+      "loss": 1.4021,
+      "step": 7980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9836009740829468,
+      "learning_rate": 4.789294210943695e-05,
+      "loss": 1.5316,
+      "step": 7981
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5112876892089844,
+      "learning_rate": 4.7892677768966434e-05,
+      "loss": 1.3713,
+      "step": 7982
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6534385681152344,
+      "learning_rate": 4.78924134284959e-05,
+      "loss": 1.3616,
+      "step": 7983
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5049601793289185,
+      "learning_rate": 4.789214908802538e-05,
+      "loss": 1.3898,
+      "step": 7984
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.613889455795288,
+      "learning_rate": 4.7891884747554854e-05,
+      "loss": 1.3635,
+      "step": 7985
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7444796562194824,
+      "learning_rate": 4.789162040708433e-05,
+      "loss": 1.4273,
+      "step": 7986
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6494979858398438,
+      "learning_rate": 4.78913560666138e-05,
+      "loss": 1.2991,
+      "step": 7987
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6599719524383545,
+      "learning_rate": 4.7891091726143275e-05,
+      "loss": 1.5131,
+      "step": 7988
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9226008653640747,
+      "learning_rate": 4.789082738567275e-05,
+      "loss": 1.4047,
+      "step": 7989
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7422367334365845,
+      "learning_rate": 4.789056304520222e-05,
+      "loss": 1.314,
+      "step": 7990
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7277288436889648,
+      "learning_rate": 4.7890298704731696e-05,
+      "loss": 2.0894,
+      "step": 7991
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.567671298980713,
+      "learning_rate": 4.789003436426117e-05,
+      "loss": 1.3317,
+      "step": 7992
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4427118301391602,
+      "learning_rate": 4.788977002379064e-05,
+      "loss": 1.2216,
+      "step": 7993
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5882933139801025,
+      "learning_rate": 4.7889505683320116e-05,
+      "loss": 1.23,
+      "step": 7994
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1266942024230957,
+      "learning_rate": 4.78892413428496e-05,
+      "loss": 1.5133,
+      "step": 7995
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6124635934829712,
+      "learning_rate": 4.7888977002379064e-05,
+      "loss": 1.5219,
+      "step": 7996
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.218451738357544,
+      "learning_rate": 4.7888712661908544e-05,
+      "loss": 1.3685,
+      "step": 7997
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8429006338119507,
+      "learning_rate": 4.788844832143801e-05,
+      "loss": 1.511,
+      "step": 7998
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9316781759262085,
+      "learning_rate": 4.788818398096749e-05,
+      "loss": 1.2304,
+      "step": 7999
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5199123620986938,
+      "learning_rate": 4.7887919640496965e-05,
+      "loss": 1.3195,
+      "step": 8000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2073915004730225,
+      "learning_rate": 4.788765530002644e-05,
+      "loss": 1.4123,
+      "step": 8001
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9270038604736328,
+      "learning_rate": 4.788739095955591e-05,
+      "loss": 1.4505,
+      "step": 8002
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6128722429275513,
+      "learning_rate": 4.788712661908538e-05,
+      "loss": 1.2904,
+      "step": 8003
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4177501201629639,
+      "learning_rate": 4.788686227861486e-05,
+      "loss": 1.2052,
+      "step": 8004
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7820614576339722,
+      "learning_rate": 4.788659793814433e-05,
+      "loss": 1.362,
+      "step": 8005
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9763988256454468,
+      "learning_rate": 4.7886333597673806e-05,
+      "loss": 1.4666,
+      "step": 8006
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.699318766593933,
+      "learning_rate": 4.788606925720328e-05,
+      "loss": 1.4662,
+      "step": 8007
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1298410892486572,
+      "learning_rate": 4.788580491673275e-05,
+      "loss": 1.506,
+      "step": 8008
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.592982292175293,
+      "learning_rate": 4.788554057626223e-05,
+      "loss": 1.3513,
+      "step": 8009
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0237677097320557,
+      "learning_rate": 4.788527623579171e-05,
+      "loss": 1.4456,
+      "step": 8010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0238888263702393,
+      "learning_rate": 4.7885011895321174e-05,
+      "loss": 1.5435,
+      "step": 8011
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6276954412460327,
+      "learning_rate": 4.788474755485065e-05,
+      "loss": 1.4833,
+      "step": 8012
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.17417311668396,
+      "learning_rate": 4.788448321438012e-05,
+      "loss": 1.3686,
+      "step": 8013
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7170652151107788,
+      "learning_rate": 4.7884218873909595e-05,
+      "loss": 1.4132,
+      "step": 8014
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.023829698562622,
+      "learning_rate": 4.7883954533439075e-05,
+      "loss": 1.3339,
+      "step": 8015
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9040875434875488,
+      "learning_rate": 4.788369019296854e-05,
+      "loss": 1.4467,
+      "step": 8016
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7354300022125244,
+      "learning_rate": 4.788342585249802e-05,
+      "loss": 1.2875,
+      "step": 8017
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6504930257797241,
+      "learning_rate": 4.788316151202749e-05,
+      "loss": 1.3951,
+      "step": 8018
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0703370571136475,
+      "learning_rate": 4.788289717155697e-05,
+      "loss": 1.2576,
+      "step": 8019
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5210199356079102,
+      "learning_rate": 4.788263283108644e-05,
+      "loss": 1.3002,
+      "step": 8020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8080960512161255,
+      "learning_rate": 4.7882368490615916e-05,
+      "loss": 1.422,
+      "step": 8021
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0381054878234863,
+      "learning_rate": 4.788210415014539e-05,
+      "loss": 1.471,
+      "step": 8022
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6348202228546143,
+      "learning_rate": 4.7881839809674864e-05,
+      "loss": 2.3391,
+      "step": 8023
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8573423624038696,
+      "learning_rate": 4.788157546920434e-05,
+      "loss": 1.4498,
+      "step": 8024
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1143100261688232,
+      "learning_rate": 4.788131112873381e-05,
+      "loss": 1.3713,
+      "step": 8025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.497302770614624,
+      "learning_rate": 4.7881046788263284e-05,
+      "loss": 1.4382,
+      "step": 8026
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9508190155029297,
+      "learning_rate": 4.788078244779276e-05,
+      "loss": 1.3539,
+      "step": 8027
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8844938278198242,
+      "learning_rate": 4.788051810732223e-05,
+      "loss": 1.5361,
+      "step": 8028
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.032240390777588,
+      "learning_rate": 4.7880253766851705e-05,
+      "loss": 1.4013,
+      "step": 8029
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8818771839141846,
+      "learning_rate": 4.7879989426381185e-05,
+      "loss": 1.4512,
+      "step": 8030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6781299114227295,
+      "learning_rate": 4.787972508591065e-05,
+      "loss": 1.4952,
+      "step": 8031
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5461301803588867,
+      "learning_rate": 4.787946074544013e-05,
+      "loss": 1.4084,
+      "step": 8032
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5309147834777832,
+      "learning_rate": 4.78791964049696e-05,
+      "loss": 2.1824,
+      "step": 8033
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0274181365966797,
+      "learning_rate": 4.787893206449908e-05,
+      "loss": 1.7001,
+      "step": 8034
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5764105319976807,
+      "learning_rate": 4.787866772402855e-05,
+      "loss": 1.4128,
+      "step": 8035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8921728134155273,
+      "learning_rate": 4.787840338355803e-05,
+      "loss": 1.3066,
+      "step": 8036
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5911121368408203,
+      "learning_rate": 4.78781390430875e-05,
+      "loss": 1.4086,
+      "step": 8037
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.623842477798462,
+      "learning_rate": 4.787787470261697e-05,
+      "loss": 1.3534,
+      "step": 8038
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5071601867675781,
+      "learning_rate": 4.787761036214645e-05,
+      "loss": 1.2219,
+      "step": 8039
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.109330415725708,
+      "learning_rate": 4.787734602167592e-05,
+      "loss": 1.4712,
+      "step": 8040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1747233867645264,
+      "learning_rate": 4.7877081681205395e-05,
+      "loss": 1.4801,
+      "step": 8041
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6434210538864136,
+      "learning_rate": 4.787681734073487e-05,
+      "loss": 1.3637,
+      "step": 8042
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5585802793502808,
+      "learning_rate": 4.787655300026434e-05,
+      "loss": 1.3641,
+      "step": 8043
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7821091413497925,
+      "learning_rate": 4.7876288659793815e-05,
+      "loss": 1.4224,
+      "step": 8044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7667971849441528,
+      "learning_rate": 4.7876024319323296e-05,
+      "loss": 1.2562,
+      "step": 8045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8418461084365845,
+      "learning_rate": 4.787575997885276e-05,
+      "loss": 1.3223,
+      "step": 8046
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.763735294342041,
+      "learning_rate": 4.787549563838224e-05,
+      "loss": 1.4158,
+      "step": 8047
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2643836736679077,
+      "learning_rate": 4.787523129791171e-05,
+      "loss": 1.19,
+      "step": 8048
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.916824460029602,
+      "learning_rate": 4.787496695744118e-05,
+      "loss": 1.4058,
+      "step": 8049
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0144433975219727,
+      "learning_rate": 4.7874702616970664e-05,
+      "loss": 1.3955,
+      "step": 8050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2145755290985107,
+      "learning_rate": 4.787443827650013e-05,
+      "loss": 1.3802,
+      "step": 8051
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6471946239471436,
+      "learning_rate": 4.787417393602961e-05,
+      "loss": 1.2321,
+      "step": 8052
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.082120180130005,
+      "learning_rate": 4.787390959555908e-05,
+      "loss": 1.4612,
+      "step": 8053
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4132866859436035,
+      "learning_rate": 4.787364525508856e-05,
+      "loss": 1.9531,
+      "step": 8054
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9274383783340454,
+      "learning_rate": 4.787338091461803e-05,
+      "loss": 1.4348,
+      "step": 8055
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6493961811065674,
+      "learning_rate": 4.7873116574147505e-05,
+      "loss": 1.4618,
+      "step": 8056
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8311495780944824,
+      "learning_rate": 4.787285223367698e-05,
+      "loss": 1.5334,
+      "step": 8057
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8400235176086426,
+      "learning_rate": 4.787258789320645e-05,
+      "loss": 1.5047,
+      "step": 8058
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5276658535003662,
+      "learning_rate": 4.7872323552735926e-05,
+      "loss": 1.3242,
+      "step": 8059
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8507503271102905,
+      "learning_rate": 4.78720592122654e-05,
+      "loss": 1.3966,
+      "step": 8060
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3620105981826782,
+      "learning_rate": 4.787179487179487e-05,
+      "loss": 1.5855,
+      "step": 8061
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6126271486282349,
+      "learning_rate": 4.7871530531324346e-05,
+      "loss": 1.3158,
+      "step": 8062
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.755212664604187,
+      "learning_rate": 4.787126619085382e-05,
+      "loss": 1.3632,
+      "step": 8063
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0969467163085938,
+      "learning_rate": 4.7871001850383294e-05,
+      "loss": 1.4029,
+      "step": 8064
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9337419271469116,
+      "learning_rate": 4.7870737509912774e-05,
+      "loss": 1.6687,
+      "step": 8065
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9784862995147705,
+      "learning_rate": 4.787047316944224e-05,
+      "loss": 1.5169,
+      "step": 8066
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7805384397506714,
+      "learning_rate": 4.787020882897172e-05,
+      "loss": 1.5551,
+      "step": 8067
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.078350067138672,
+      "learning_rate": 4.786994448850119e-05,
+      "loss": 1.4243,
+      "step": 8068
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7141664028167725,
+      "learning_rate": 4.786968014803067e-05,
+      "loss": 1.3784,
+      "step": 8069
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7348814010620117,
+      "learning_rate": 4.786941580756014e-05,
+      "loss": 1.4676,
+      "step": 8070
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7003589868545532,
+      "learning_rate": 4.7869151467089615e-05,
+      "loss": 1.333,
+      "step": 8071
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.734415054321289,
+      "learning_rate": 4.786888712661909e-05,
+      "loss": 2.3874,
+      "step": 8072
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6282938718795776,
+      "learning_rate": 4.786862278614856e-05,
+      "loss": 1.2905,
+      "step": 8073
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2723265886306763,
+      "learning_rate": 4.7868358445678036e-05,
+      "loss": 1.9753,
+      "step": 8074
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5665557384490967,
+      "learning_rate": 4.786809410520751e-05,
+      "loss": 1.2777,
+      "step": 8075
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8464711904525757,
+      "learning_rate": 4.786782976473698e-05,
+      "loss": 1.376,
+      "step": 8076
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4633594751358032,
+      "learning_rate": 4.786756542426646e-05,
+      "loss": 1.3616,
+      "step": 8077
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.585845708847046,
+      "learning_rate": 4.786730108379593e-05,
+      "loss": 1.2451,
+      "step": 8078
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8175015449523926,
+      "learning_rate": 4.7867036743325404e-05,
+      "loss": 1.5778,
+      "step": 8079
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.498710036277771,
+      "learning_rate": 4.7866772402854884e-05,
+      "loss": 1.3345,
+      "step": 8080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9620070457458496,
+      "learning_rate": 4.786650806238435e-05,
+      "loss": 1.4487,
+      "step": 8081
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3090068101882935,
+      "learning_rate": 4.786624372191383e-05,
+      "loss": 1.2824,
+      "step": 8082
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7251641750335693,
+      "learning_rate": 4.78659793814433e-05,
+      "loss": 1.2911,
+      "step": 8083
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.434356689453125,
+      "learning_rate": 4.786571504097277e-05,
+      "loss": 1.349,
+      "step": 8084
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5810909271240234,
+      "learning_rate": 4.786545070050225e-05,
+      "loss": 1.437,
+      "step": 8085
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7287013530731201,
+      "learning_rate": 4.786518636003172e-05,
+      "loss": 1.3048,
+      "step": 8086
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.054036855697632,
+      "learning_rate": 4.78649220195612e-05,
+      "loss": 1.3625,
+      "step": 8087
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.623754620552063,
+      "learning_rate": 4.7864657679090666e-05,
+      "loss": 1.473,
+      "step": 8088
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6355037689208984,
+      "learning_rate": 4.7864393338620146e-05,
+      "loss": 1.2664,
+      "step": 8089
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5994571447372437,
+      "learning_rate": 4.786412899814962e-05,
+      "loss": 1.4164,
+      "step": 8090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.006045341491699,
+      "learning_rate": 4.7863864657679094e-05,
+      "loss": 1.4011,
+      "step": 8091
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6260695457458496,
+      "learning_rate": 4.786360031720857e-05,
+      "loss": 1.4047,
+      "step": 8092
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0591607093811035,
+      "learning_rate": 4.786333597673804e-05,
+      "loss": 1.4319,
+      "step": 8093
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.509019136428833,
+      "learning_rate": 4.7863071636267514e-05,
+      "loss": 1.3749,
+      "step": 8094
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7695242166519165,
+      "learning_rate": 4.786280729579699e-05,
+      "loss": 1.4166,
+      "step": 8095
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.686521053314209,
+      "learning_rate": 4.786254295532646e-05,
+      "loss": 1.4182,
+      "step": 8096
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7946640253067017,
+      "learning_rate": 4.7862278614855935e-05,
+      "loss": 1.2493,
+      "step": 8097
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5239092111587524,
+      "learning_rate": 4.786201427438541e-05,
+      "loss": 1.3229,
+      "step": 8098
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8289928436279297,
+      "learning_rate": 4.786174993391488e-05,
+      "loss": 1.4106,
+      "step": 8099
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7424969673156738,
+      "learning_rate": 4.786148559344436e-05,
+      "loss": 1.2855,
+      "step": 8100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8988800048828125,
+      "learning_rate": 4.786122125297383e-05,
+      "loss": 1.321,
+      "step": 8101
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6902433633804321,
+      "learning_rate": 4.786095691250331e-05,
+      "loss": 1.3408,
+      "step": 8102
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.668629765510559,
+      "learning_rate": 4.7860692572032776e-05,
+      "loss": 1.4644,
+      "step": 8103
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9416004419326782,
+      "learning_rate": 4.786042823156226e-05,
+      "loss": 1.4462,
+      "step": 8104
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5879853963851929,
+      "learning_rate": 4.786016389109173e-05,
+      "loss": 1.4405,
+      "step": 8105
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7088550329208374,
+      "learning_rate": 4.7859899550621204e-05,
+      "loss": 1.3298,
+      "step": 8106
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6209111213684082,
+      "learning_rate": 4.785963521015068e-05,
+      "loss": 1.3974,
+      "step": 8107
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.435187578201294,
+      "learning_rate": 4.785937086968015e-05,
+      "loss": 1.287,
+      "step": 8108
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.216527223587036,
+      "learning_rate": 4.7859106529209625e-05,
+      "loss": 1.3879,
+      "step": 8109
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6762299537658691,
+      "learning_rate": 4.78588421887391e-05,
+      "loss": 1.6588,
+      "step": 8110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9839978218078613,
+      "learning_rate": 4.785857784826857e-05,
+      "loss": 2.1617,
+      "step": 8111
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8081517219543457,
+      "learning_rate": 4.7858313507798045e-05,
+      "loss": 1.337,
+      "step": 8112
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7576344013214111,
+      "learning_rate": 4.785804916732752e-05,
+      "loss": 1.3312,
+      "step": 8113
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5538127422332764,
+      "learning_rate": 4.785778482685699e-05,
+      "loss": 1.3093,
+      "step": 8114
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2328664064407349,
+      "learning_rate": 4.785752048638647e-05,
+      "loss": 1.2149,
+      "step": 8115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.285789966583252,
+      "learning_rate": 4.785725614591594e-05,
+      "loss": 1.612,
+      "step": 8116
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.597400188446045,
+      "learning_rate": 4.785699180544542e-05,
+      "loss": 1.3791,
+      "step": 8117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6221139430999756,
+      "learning_rate": 4.785672746497489e-05,
+      "loss": 1.1122,
+      "step": 8118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8140006065368652,
+      "learning_rate": 4.785646312450437e-05,
+      "loss": 1.401,
+      "step": 8119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.782577633857727,
+      "learning_rate": 4.785619878403384e-05,
+      "loss": 1.3306,
+      "step": 8120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.944520354270935,
+      "learning_rate": 4.785593444356331e-05,
+      "loss": 1.4008,
+      "step": 8121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7930258512496948,
+      "learning_rate": 4.785567010309279e-05,
+      "loss": 1.4392,
+      "step": 8122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4712704420089722,
+      "learning_rate": 4.7855405762622255e-05,
+      "loss": 1.3511,
+      "step": 8123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5404740571975708,
+      "learning_rate": 4.7855141422151735e-05,
+      "loss": 1.2166,
+      "step": 8124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5294759273529053,
+      "learning_rate": 4.785487708168121e-05,
+      "loss": 1.3252,
+      "step": 8125
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5707587003707886,
+      "learning_rate": 4.785461274121068e-05,
+      "loss": 1.4055,
+      "step": 8126
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.838070273399353,
+      "learning_rate": 4.7854348400740156e-05,
+      "loss": 1.3372,
+      "step": 8127
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4866551160812378,
+      "learning_rate": 4.785408406026963e-05,
+      "loss": 1.3639,
+      "step": 8128
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5165798664093018,
+      "learning_rate": 4.78538197197991e-05,
+      "loss": 1.4587,
+      "step": 8129
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6281243562698364,
+      "learning_rate": 4.7853555379328576e-05,
+      "loss": 1.2324,
+      "step": 8130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.519271969795227,
+      "learning_rate": 4.785329103885805e-05,
+      "loss": 1.4561,
+      "step": 8131
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.499821901321411,
+      "learning_rate": 4.7853026698387523e-05,
+      "loss": 1.6433,
+      "step": 8132
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8493599891662598,
+      "learning_rate": 4.7852762357917e-05,
+      "loss": 1.2589,
+      "step": 8133
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2492268085479736,
+      "learning_rate": 4.785249801744647e-05,
+      "loss": 1.5429,
+      "step": 8134
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5565024614334106,
+      "learning_rate": 4.785223367697595e-05,
+      "loss": 1.2921,
+      "step": 8135
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8827863931655884,
+      "learning_rate": 4.785196933650542e-05,
+      "loss": 1.4662,
+      "step": 8136
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1091647148132324,
+      "learning_rate": 4.78517049960349e-05,
+      "loss": 1.5592,
+      "step": 8137
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.666088581085205,
+      "learning_rate": 4.7851440655564365e-05,
+      "loss": 1.4558,
+      "step": 8138
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1348860263824463,
+      "learning_rate": 4.7851176315093845e-05,
+      "loss": 1.5662,
+      "step": 8139
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.545863151550293,
+      "learning_rate": 4.785091197462332e-05,
+      "loss": 1.5051,
+      "step": 8140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7149226665496826,
+      "learning_rate": 4.785064763415279e-05,
+      "loss": 1.3488,
+      "step": 8141
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6289163827896118,
+      "learning_rate": 4.7850383293682266e-05,
+      "loss": 1.3908,
+      "step": 8142
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4387091398239136,
+      "learning_rate": 4.785011895321174e-05,
+      "loss": 1.4554,
+      "step": 8143
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.308249592781067,
+      "learning_rate": 4.784985461274121e-05,
+      "loss": 1.998,
+      "step": 8144
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7999932765960693,
+      "learning_rate": 4.784959027227069e-05,
+      "loss": 1.38,
+      "step": 8145
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4523024559020996,
+      "learning_rate": 4.784932593180016e-05,
+      "loss": 1.2448,
+      "step": 8146
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.791330099105835,
+      "learning_rate": 4.7849061591329634e-05,
+      "loss": 1.4353,
+      "step": 8147
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.899702787399292,
+      "learning_rate": 4.784879725085911e-05,
+      "loss": 1.6187,
+      "step": 8148
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7528084516525269,
+      "learning_rate": 4.784853291038858e-05,
+      "loss": 1.5751,
+      "step": 8149
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9242067337036133,
+      "learning_rate": 4.784826856991806e-05,
+      "loss": 1.537,
+      "step": 8150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.232008218765259,
+      "learning_rate": 4.784800422944753e-05,
+      "loss": 1.5242,
+      "step": 8151
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5389974117279053,
+      "learning_rate": 4.784773988897701e-05,
+      "loss": 2.1743,
+      "step": 8152
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5566139221191406,
+      "learning_rate": 4.7847475548506475e-05,
+      "loss": 1.9853,
+      "step": 8153
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6741645336151123,
+      "learning_rate": 4.7847211208035956e-05,
+      "loss": 1.2605,
+      "step": 8154
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.391153335571289,
+      "learning_rate": 4.784694686756543e-05,
+      "loss": 1.3487,
+      "step": 8155
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8559246063232422,
+      "learning_rate": 4.7846682527094896e-05,
+      "loss": 1.5432,
+      "step": 8156
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8355623483657837,
+      "learning_rate": 4.7846418186624376e-05,
+      "loss": 1.3594,
+      "step": 8157
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.594419002532959,
+      "learning_rate": 4.784615384615384e-05,
+      "loss": 1.3651,
+      "step": 8158
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7071269750595093,
+      "learning_rate": 4.7845889505683323e-05,
+      "loss": 1.3777,
+      "step": 8159
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3789124488830566,
+      "learning_rate": 4.78456251652128e-05,
+      "loss": 1.6874,
+      "step": 8160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7127902507781982,
+      "learning_rate": 4.784536082474227e-05,
+      "loss": 1.3444,
+      "step": 8161
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.60786771774292,
+      "learning_rate": 4.7845096484271744e-05,
+      "loss": 1.4619,
+      "step": 8162
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8947690725326538,
+      "learning_rate": 4.784483214380122e-05,
+      "loss": 1.371,
+      "step": 8163
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.341980457305908,
+      "learning_rate": 4.784456780333069e-05,
+      "loss": 1.5712,
+      "step": 8164
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8241020441055298,
+      "learning_rate": 4.784430346286017e-05,
+      "loss": 1.4257,
+      "step": 8165
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5499792098999023,
+      "learning_rate": 4.784403912238964e-05,
+      "loss": 1.3284,
+      "step": 8166
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3357393741607666,
+      "learning_rate": 4.784377478191911e-05,
+      "loss": 1.5175,
+      "step": 8167
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5775688886642456,
+      "learning_rate": 4.7843510441448586e-05,
+      "loss": 1.2815,
+      "step": 8168
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.927551507949829,
+      "learning_rate": 4.784324610097806e-05,
+      "loss": 1.4227,
+      "step": 8169
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6542226076126099,
+      "learning_rate": 4.784298176050754e-05,
+      "loss": 1.3091,
+      "step": 8170
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9714992046356201,
+      "learning_rate": 4.7842717420037006e-05,
+      "loss": 1.6377,
+      "step": 8171
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9262394905090332,
+      "learning_rate": 4.784245307956649e-05,
+      "loss": 1.2906,
+      "step": 8172
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5991535186767578,
+      "learning_rate": 4.7842188739095953e-05,
+      "loss": 1.309,
+      "step": 8173
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.503429889678955,
+      "learning_rate": 4.7841924398625434e-05,
+      "loss": 2.226,
+      "step": 8174
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9620801210403442,
+      "learning_rate": 4.784166005815491e-05,
+      "loss": 1.3866,
+      "step": 8175
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2796456813812256,
+      "learning_rate": 4.784139571768438e-05,
+      "loss": 1.6666,
+      "step": 8176
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.775582194328308,
+      "learning_rate": 4.7841131377213855e-05,
+      "loss": 1.2733,
+      "step": 8177
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4298946857452393,
+      "learning_rate": 4.784086703674333e-05,
+      "loss": 1.3546,
+      "step": 8178
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7139946222305298,
+      "learning_rate": 4.78406026962728e-05,
+      "loss": 1.2804,
+      "step": 8179
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8394397497177124,
+      "learning_rate": 4.7840338355802275e-05,
+      "loss": 1.3894,
+      "step": 8180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0846073627471924,
+      "learning_rate": 4.784007401533175e-05,
+      "loss": 1.3909,
+      "step": 8181
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.740971565246582,
+      "learning_rate": 4.783980967486122e-05,
+      "loss": 1.3938,
+      "step": 8182
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.69500732421875,
+      "learning_rate": 4.7839545334390696e-05,
+      "loss": 1.3418,
+      "step": 8183
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.599096655845642,
+      "learning_rate": 4.783928099392017e-05,
+      "loss": 1.3644,
+      "step": 8184
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5828328132629395,
+      "learning_rate": 4.783901665344965e-05,
+      "loss": 1.3598,
+      "step": 8185
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7586568593978882,
+      "learning_rate": 4.783875231297912e-05,
+      "loss": 1.3484,
+      "step": 8186
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9984582662582397,
+      "learning_rate": 4.78384879725086e-05,
+      "loss": 1.3443,
+      "step": 8187
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1308958530426025,
+      "learning_rate": 4.7838223632038064e-05,
+      "loss": 1.5363,
+      "step": 8188
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.621407151222229,
+      "learning_rate": 4.7837959291567544e-05,
+      "loss": 1.4392,
+      "step": 8189
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7967172861099243,
+      "learning_rate": 4.783769495109702e-05,
+      "loss": 1.4693,
+      "step": 8190
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4360649585723877,
+      "learning_rate": 4.783743061062649e-05,
+      "loss": 1.4299,
+      "step": 8191
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.953202247619629,
+      "learning_rate": 4.7837166270155965e-05,
+      "loss": 1.5469,
+      "step": 8192
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5180801153182983,
+      "learning_rate": 4.783690192968543e-05,
+      "loss": 1.3744,
+      "step": 8193
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6166263818740845,
+      "learning_rate": 4.783663758921491e-05,
+      "loss": 1.3421,
+      "step": 8194
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8533642292022705,
+      "learning_rate": 4.7836373248744386e-05,
+      "loss": 1.3568,
+      "step": 8195
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9574389457702637,
+      "learning_rate": 4.783610890827386e-05,
+      "loss": 1.4452,
+      "step": 8196
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6141139268875122,
+      "learning_rate": 4.783584456780333e-05,
+      "loss": 1.39,
+      "step": 8197
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5159549713134766,
+      "learning_rate": 4.7835580227332806e-05,
+      "loss": 1.5349,
+      "step": 8198
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4335389137268066,
+      "learning_rate": 4.783531588686228e-05,
+      "loss": 1.3146,
+      "step": 8199
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.720055341720581,
+      "learning_rate": 4.783505154639176e-05,
+      "loss": 1.8285,
+      "step": 8200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6672868728637695,
+      "learning_rate": 4.783478720592123e-05,
+      "loss": 1.3137,
+      "step": 8201
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6669646501541138,
+      "learning_rate": 4.78345228654507e-05,
+      "loss": 1.3884,
+      "step": 8202
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4826210737228394,
+      "learning_rate": 4.7834258524980174e-05,
+      "loss": 1.4357,
+      "step": 8203
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0956761837005615,
+      "learning_rate": 4.783399418450965e-05,
+      "loss": 1.3083,
+      "step": 8204
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8581727743148804,
+      "learning_rate": 4.783372984403913e-05,
+      "loss": 1.4546,
+      "step": 8205
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6795555353164673,
+      "learning_rate": 4.7833465503568595e-05,
+      "loss": 1.414,
+      "step": 8206
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3636332750320435,
+      "learning_rate": 4.7833201163098075e-05,
+      "loss": 1.3995,
+      "step": 8207
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7439550161361694,
+      "learning_rate": 4.783293682262754e-05,
+      "loss": 1.423,
+      "step": 8208
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.738818645477295,
+      "learning_rate": 4.783267248215702e-05,
+      "loss": 1.6949,
+      "step": 8209
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9127209186553955,
+      "learning_rate": 4.7832408141686496e-05,
+      "loss": 1.4408,
+      "step": 8210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.149202823638916,
+      "learning_rate": 4.783214380121597e-05,
+      "loss": 1.6731,
+      "step": 8211
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8637748956680298,
+      "learning_rate": 4.783187946074544e-05,
+      "loss": 1.2905,
+      "step": 8212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.546596884727478,
+      "learning_rate": 4.783161512027492e-05,
+      "loss": 1.4475,
+      "step": 8213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8345887660980225,
+      "learning_rate": 4.783135077980439e-05,
+      "loss": 1.6176,
+      "step": 8214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8262442350387573,
+      "learning_rate": 4.7831086439333864e-05,
+      "loss": 1.2727,
+      "step": 8215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6839426755905151,
+      "learning_rate": 4.783082209886334e-05,
+      "loss": 1.3476,
+      "step": 8216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8288434743881226,
+      "learning_rate": 4.783055775839281e-05,
+      "loss": 1.4588,
+      "step": 8217
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7754831314086914,
+      "learning_rate": 4.7830293417922284e-05,
+      "loss": 1.514,
+      "step": 8218
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7275817394256592,
+      "learning_rate": 4.783002907745176e-05,
+      "loss": 1.4186,
+      "step": 8219
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9322744607925415,
+      "learning_rate": 4.782976473698124e-05,
+      "loss": 1.5796,
+      "step": 8220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.923931360244751,
+      "learning_rate": 4.7829500396510705e-05,
+      "loss": 1.324,
+      "step": 8221
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.747785806655884,
+      "learning_rate": 4.7829236056040186e-05,
+      "loss": 1.553,
+      "step": 8222
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0372307300567627,
+      "learning_rate": 4.782897171556965e-05,
+      "loss": 1.4734,
+      "step": 8223
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.754443883895874,
+      "learning_rate": 4.782870737509913e-05,
+      "loss": 1.4029,
+      "step": 8224
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.896212339401245,
+      "learning_rate": 4.7828443034628606e-05,
+      "loss": 1.525,
+      "step": 8225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3554730415344238,
+      "learning_rate": 4.782817869415808e-05,
+      "loss": 1.7596,
+      "step": 8226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8032288551330566,
+      "learning_rate": 4.782791435368755e-05,
+      "loss": 1.4409,
+      "step": 8227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.695813536643982,
+      "learning_rate": 4.782765001321702e-05,
+      "loss": 1.4024,
+      "step": 8228
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.483328104019165,
+      "learning_rate": 4.78273856727465e-05,
+      "loss": 1.4495,
+      "step": 8229
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.697622299194336,
+      "learning_rate": 4.7827121332275974e-05,
+      "loss": 1.2901,
+      "step": 8230
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8738504648208618,
+      "learning_rate": 4.782685699180545e-05,
+      "loss": 1.5043,
+      "step": 8231
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7039599418640137,
+      "learning_rate": 4.782659265133492e-05,
+      "loss": 1.3063,
+      "step": 8232
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4965780973434448,
+      "learning_rate": 4.7826328310864395e-05,
+      "loss": 1.3273,
+      "step": 8233
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1757566928863525,
+      "learning_rate": 4.782606397039387e-05,
+      "loss": 1.5375,
+      "step": 8234
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.515748143196106,
+      "learning_rate": 4.782579962992335e-05,
+      "loss": 1.4912,
+      "step": 8235
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.373883605003357,
+      "learning_rate": 4.7825535289452816e-05,
+      "loss": 1.3293,
+      "step": 8236
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.020514726638794,
+      "learning_rate": 4.7825270948982296e-05,
+      "loss": 1.482,
+      "step": 8237
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.594559669494629,
+      "learning_rate": 4.782500660851176e-05,
+      "loss": 1.4325,
+      "step": 8238
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.929203987121582,
+      "learning_rate": 4.7824742268041236e-05,
+      "loss": 1.4913,
+      "step": 8239
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.130296468734741,
+      "learning_rate": 4.7824477927570717e-05,
+      "loss": 1.4397,
+      "step": 8240
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.410520076751709,
+      "learning_rate": 4.782421358710018e-05,
+      "loss": 1.3959,
+      "step": 8241
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.316987991333008,
+      "learning_rate": 4.7823949246629664e-05,
+      "loss": 3.3223,
+      "step": 8242
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.644054651260376,
+      "learning_rate": 4.782368490615913e-05,
+      "loss": 1.8161,
+      "step": 8243
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5212668180465698,
+      "learning_rate": 4.782342056568861e-05,
+      "loss": 1.3917,
+      "step": 8244
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8903965950012207,
+      "learning_rate": 4.7823156225218084e-05,
+      "loss": 1.4669,
+      "step": 8245
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1150710582733154,
+      "learning_rate": 4.782289188474756e-05,
+      "loss": 1.4796,
+      "step": 8246
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.841048002243042,
+      "learning_rate": 4.782262754427703e-05,
+      "loss": 1.3722,
+      "step": 8247
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5086188316345215,
+      "learning_rate": 4.7822363203806505e-05,
+      "loss": 1.3579,
+      "step": 8248
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5203665494918823,
+      "learning_rate": 4.782209886333598e-05,
+      "loss": 1.4927,
+      "step": 8249
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4539580345153809,
+      "learning_rate": 4.782183452286545e-05,
+      "loss": 1.2382,
+      "step": 8250
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.317018747329712,
+      "learning_rate": 4.7821570182394926e-05,
+      "loss": 1.3081,
+      "step": 8251
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5662802457809448,
+      "learning_rate": 4.78213058419244e-05,
+      "loss": 1.3995,
+      "step": 8252
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1422922611236572,
+      "learning_rate": 4.782104150145387e-05,
+      "loss": 1.4644,
+      "step": 8253
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4391486644744873,
+      "learning_rate": 4.7820777160983347e-05,
+      "loss": 1.6443,
+      "step": 8254
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5960885286331177,
+      "learning_rate": 4.782051282051283e-05,
+      "loss": 1.3055,
+      "step": 8255
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9956756830215454,
+      "learning_rate": 4.7820248480042294e-05,
+      "loss": 1.4592,
+      "step": 8256
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3126754760742188,
+      "learning_rate": 4.7819984139571774e-05,
+      "loss": 2.3587,
+      "step": 8257
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.150838851928711,
+      "learning_rate": 4.781971979910124e-05,
+      "loss": 1.4339,
+      "step": 8258
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.052324056625366,
+      "learning_rate": 4.781945545863072e-05,
+      "loss": 1.3989,
+      "step": 8259
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8393714427947998,
+      "learning_rate": 4.7819191118160195e-05,
+      "loss": 1.3936,
+      "step": 8260
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2479469776153564,
+      "learning_rate": 4.781892677768967e-05,
+      "loss": 1.3883,
+      "step": 8261
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.844332218170166,
+      "learning_rate": 4.781866243721914e-05,
+      "loss": 1.5276,
+      "step": 8262
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2058327198028564,
+      "learning_rate": 4.7818398096748615e-05,
+      "loss": 1.4917,
+      "step": 8263
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.848699688911438,
+      "learning_rate": 4.781813375627809e-05,
+      "loss": 1.396,
+      "step": 8264
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5665770769119263,
+      "learning_rate": 4.781786941580756e-05,
+      "loss": 1.3958,
+      "step": 8265
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8261418342590332,
+      "learning_rate": 4.7817605075337036e-05,
+      "loss": 1.4416,
+      "step": 8266
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7149486541748047,
+      "learning_rate": 4.781734073486651e-05,
+      "loss": 1.4701,
+      "step": 8267
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8275874853134155,
+      "learning_rate": 4.781707639439598e-05,
+      "loss": 1.5472,
+      "step": 8268
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3937463760375977,
+      "learning_rate": 4.781681205392546e-05,
+      "loss": 1.6738,
+      "step": 8269
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9682189226150513,
+      "learning_rate": 4.781654771345494e-05,
+      "loss": 1.3816,
+      "step": 8270
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8109033107757568,
+      "learning_rate": 4.7816283372984404e-05,
+      "loss": 1.3927,
+      "step": 8271
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2821478843688965,
+      "learning_rate": 4.7816019032513884e-05,
+      "loss": 1.3776,
+      "step": 8272
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7810864448547363,
+      "learning_rate": 4.781575469204335e-05,
+      "loss": 1.4188,
+      "step": 8273
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7861047983169556,
+      "learning_rate": 4.7815490351572825e-05,
+      "loss": 1.4875,
+      "step": 8274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.662398338317871,
+      "learning_rate": 4.7815226011102305e-05,
+      "loss": 1.3694,
+      "step": 8275
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8064112663269043,
+      "learning_rate": 4.781496167063177e-05,
+      "loss": 1.4243,
+      "step": 8276
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0878336429595947,
+      "learning_rate": 4.781469733016125e-05,
+      "loss": 1.6272,
+      "step": 8277
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7044388055801392,
+      "learning_rate": 4.781443298969072e-05,
+      "loss": 1.3694,
+      "step": 8278
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7075711488723755,
+      "learning_rate": 4.78141686492202e-05,
+      "loss": 1.4242,
+      "step": 8279
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.046424388885498,
+      "learning_rate": 4.781390430874967e-05,
+      "loss": 1.669,
+      "step": 8280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3645503520965576,
+      "learning_rate": 4.7813639968279147e-05,
+      "loss": 1.4689,
+      "step": 8281
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.847152590751648,
+      "learning_rate": 4.781337562780862e-05,
+      "loss": 1.4824,
+      "step": 8282
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8317421674728394,
+      "learning_rate": 4.7813111287338094e-05,
+      "loss": 1.4079,
+      "step": 8283
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6682343482971191,
+      "learning_rate": 4.781284694686757e-05,
+      "loss": 1.2919,
+      "step": 8284
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5733942985534668,
+      "learning_rate": 4.781258260639704e-05,
+      "loss": 1.4985,
+      "step": 8285
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.286391019821167,
+      "learning_rate": 4.7812318265926514e-05,
+      "loss": 1.5849,
+      "step": 8286
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7549430131912231,
+      "learning_rate": 4.781205392545599e-05,
+      "loss": 1.4145,
+      "step": 8287
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8789736032485962,
+      "learning_rate": 4.781178958498546e-05,
+      "loss": 1.421,
+      "step": 8288
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7502902746200562,
+      "learning_rate": 4.7811525244514935e-05,
+      "loss": 1.3941,
+      "step": 8289
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.710852861404419,
+      "learning_rate": 4.7811260904044415e-05,
+      "loss": 1.478,
+      "step": 8290
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.545762062072754,
+      "learning_rate": 4.781099656357388e-05,
+      "loss": 1.1653,
+      "step": 8291
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7870495319366455,
+      "learning_rate": 4.781073222310336e-05,
+      "loss": 1.377,
+      "step": 8292
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.617181420326233,
+      "learning_rate": 4.781046788263283e-05,
+      "loss": 1.2872,
+      "step": 8293
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9826924800872803,
+      "learning_rate": 4.781020354216231e-05,
+      "loss": 1.4578,
+      "step": 8294
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1425163745880127,
+      "learning_rate": 4.780993920169178e-05,
+      "loss": 1.3446,
+      "step": 8295
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.949704647064209,
+      "learning_rate": 4.780967486122126e-05,
+      "loss": 1.4787,
+      "step": 8296
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5618231296539307,
+      "learning_rate": 4.780941052075073e-05,
+      "loss": 1.396,
+      "step": 8297
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.486501693725586,
+      "learning_rate": 4.7809146180280204e-05,
+      "loss": 1.4873,
+      "step": 8298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7571079730987549,
+      "learning_rate": 4.780888183980968e-05,
+      "loss": 1.44,
+      "step": 8299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.086930751800537,
+      "learning_rate": 4.780861749933915e-05,
+      "loss": 1.356,
+      "step": 8300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.293030023574829,
+      "learning_rate": 4.7808353158868625e-05,
+      "loss": 1.456,
+      "step": 8301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4434956312179565,
+      "learning_rate": 4.78080888183981e-05,
+      "loss": 1.3154,
+      "step": 8302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7017394304275513,
+      "learning_rate": 4.780782447792757e-05,
+      "loss": 1.4212,
+      "step": 8303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9871180057525635,
+      "learning_rate": 4.7807560137457045e-05,
+      "loss": 1.2996,
+      "step": 8304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.184863328933716,
+      "learning_rate": 4.7807295796986526e-05,
+      "loss": 1.3878,
+      "step": 8305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7167505025863647,
+      "learning_rate": 4.780703145651599e-05,
+      "loss": 1.3393,
+      "step": 8306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6457868814468384,
+      "learning_rate": 4.780676711604547e-05,
+      "loss": 1.3907,
+      "step": 8307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.823334813117981,
+      "learning_rate": 4.780650277557494e-05,
+      "loss": 1.4088,
+      "step": 8308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2095248699188232,
+      "learning_rate": 4.780623843510442e-05,
+      "loss": 1.5878,
+      "step": 8309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.473057985305786,
+      "learning_rate": 4.7805974094633894e-05,
+      "loss": 1.5893,
+      "step": 8310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8326404094696045,
+      "learning_rate": 4.780570975416336e-05,
+      "loss": 1.4073,
+      "step": 8311
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5919214487075806,
+      "learning_rate": 4.780544541369284e-05,
+      "loss": 1.2689,
+      "step": 8312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5459269285202026,
+      "learning_rate": 4.780518107322231e-05,
+      "loss": 1.3585,
+      "step": 8313
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.467873454093933,
+      "learning_rate": 4.780491673275179e-05,
+      "loss": 1.2553,
+      "step": 8314
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8026286363601685,
+      "learning_rate": 4.780465239228126e-05,
+      "loss": 1.4361,
+      "step": 8315
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9802042245864868,
+      "learning_rate": 4.7804388051810735e-05,
+      "loss": 1.2304,
+      "step": 8316
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.537078619003296,
+      "learning_rate": 4.780412371134021e-05,
+      "loss": 1.3076,
+      "step": 8317
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1042282581329346,
+      "learning_rate": 4.780385937086968e-05,
+      "loss": 1.3326,
+      "step": 8318
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6563352346420288,
+      "learning_rate": 4.7803595030399156e-05,
+      "loss": 1.3789,
+      "step": 8319
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1331560611724854,
+      "learning_rate": 4.780333068992863e-05,
+      "loss": 1.3626,
+      "step": 8320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7055397033691406,
+      "learning_rate": 4.78030663494581e-05,
+      "loss": 2.1476,
+      "step": 8321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5490169525146484,
+      "learning_rate": 4.7802802008987577e-05,
+      "loss": 1.4646,
+      "step": 8322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5678834915161133,
+      "learning_rate": 4.780253766851705e-05,
+      "loss": 1.3976,
+      "step": 8323
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9436882734298706,
+      "learning_rate": 4.7802273328046524e-05,
+      "loss": 1.4863,
+      "step": 8324
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7429418563842773,
+      "learning_rate": 4.7802008987576004e-05,
+      "loss": 1.497,
+      "step": 8325
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6572065353393555,
+      "learning_rate": 4.780174464710547e-05,
+      "loss": 1.5182,
+      "step": 8326
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7417341470718384,
+      "learning_rate": 4.780148030663495e-05,
+      "loss": 1.3249,
+      "step": 8327
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6239031553268433,
+      "learning_rate": 4.780121596616442e-05,
+      "loss": 1.424,
+      "step": 8328
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6424336433410645,
+      "learning_rate": 4.78009516256939e-05,
+      "loss": 1.4559,
+      "step": 8329
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6345841884613037,
+      "learning_rate": 4.780068728522337e-05,
+      "loss": 1.3668,
+      "step": 8330
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5239241123199463,
+      "learning_rate": 4.7800422944752845e-05,
+      "loss": 1.3942,
+      "step": 8331
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7079710960388184,
+      "learning_rate": 4.780015860428232e-05,
+      "loss": 1.517,
+      "step": 8332
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3718830347061157,
+      "learning_rate": 4.779989426381179e-05,
+      "loss": 1.6831,
+      "step": 8333
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1980860233306885,
+      "learning_rate": 4.7799629923341266e-05,
+      "loss": 1.4767,
+      "step": 8334
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5663071870803833,
+      "learning_rate": 4.779936558287074e-05,
+      "loss": 2.2318,
+      "step": 8335
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2478874921798706,
+      "learning_rate": 4.779910124240021e-05,
+      "loss": 1.1587,
+      "step": 8336
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7132066488265991,
+      "learning_rate": 4.779883690192969e-05,
+      "loss": 1.3106,
+      "step": 8337
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7805413007736206,
+      "learning_rate": 4.779857256145916e-05,
+      "loss": 1.4395,
+      "step": 8338
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.639413833618164,
+      "learning_rate": 4.7798308220988634e-05,
+      "loss": 1.4938,
+      "step": 8339
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0773913860321045,
+      "learning_rate": 4.7798043880518114e-05,
+      "loss": 1.4196,
+      "step": 8340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.189769983291626,
+      "learning_rate": 4.779777954004758e-05,
+      "loss": 1.3439,
+      "step": 8341
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2199044227600098,
+      "learning_rate": 4.779751519957706e-05,
+      "loss": 1.3966,
+      "step": 8342
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.915987491607666,
+      "learning_rate": 4.779725085910653e-05,
+      "loss": 1.4438,
+      "step": 8343
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0737385749816895,
+      "learning_rate": 4.779698651863601e-05,
+      "loss": 1.5254,
+      "step": 8344
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0282819271087646,
+      "learning_rate": 4.779672217816548e-05,
+      "loss": 1.4068,
+      "step": 8345
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.959392786026001,
+      "learning_rate": 4.779645783769495e-05,
+      "loss": 1.6332,
+      "step": 8346
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7697073221206665,
+      "learning_rate": 4.779619349722443e-05,
+      "loss": 1.4528,
+      "step": 8347
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7345832586288452,
+      "learning_rate": 4.7795929156753896e-05,
+      "loss": 1.4639,
+      "step": 8348
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9282889366149902,
+      "learning_rate": 4.7795664816283376e-05,
+      "loss": 1.4988,
+      "step": 8349
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5427134037017822,
+      "learning_rate": 4.779540047581285e-05,
+      "loss": 1.1828,
+      "step": 8350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3931750059127808,
+      "learning_rate": 4.7795136135342324e-05,
+      "loss": 1.872,
+      "step": 8351
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.548009157180786,
+      "learning_rate": 4.77948717948718e-05,
+      "loss": 1.4695,
+      "step": 8352
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5764329433441162,
+      "learning_rate": 4.779460745440127e-05,
+      "loss": 1.3617,
+      "step": 8353
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5178686380386353,
+      "learning_rate": 4.7794343113930744e-05,
+      "loss": 2.335,
+      "step": 8354
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5350782871246338,
+      "learning_rate": 4.7794078773460225e-05,
+      "loss": 1.3146,
+      "step": 8355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8243131637573242,
+      "learning_rate": 4.779381443298969e-05,
+      "loss": 1.3336,
+      "step": 8356
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.703202724456787,
+      "learning_rate": 4.7793550092519165e-05,
+      "loss": 1.3682,
+      "step": 8357
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.644543170928955,
+      "learning_rate": 4.779328575204864e-05,
+      "loss": 1.3156,
+      "step": 8358
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.468888521194458,
+      "learning_rate": 4.779302141157811e-05,
+      "loss": 1.4993,
+      "step": 8359
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.816180467605591,
+      "learning_rate": 4.779275707110759e-05,
+      "loss": 1.4461,
+      "step": 8360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.906421422958374,
+      "learning_rate": 4.779249273063706e-05,
+      "loss": 1.4922,
+      "step": 8361
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8666342496871948,
+      "learning_rate": 4.779222839016654e-05,
+      "loss": 1.4314,
+      "step": 8362
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6694514751434326,
+      "learning_rate": 4.7791964049696006e-05,
+      "loss": 1.3753,
+      "step": 8363
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9801093339920044,
+      "learning_rate": 4.779169970922549e-05,
+      "loss": 1.4105,
+      "step": 8364
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.397748351097107,
+      "learning_rate": 4.779143536875496e-05,
+      "loss": 1.3463,
+      "step": 8365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2338453531265259,
+      "learning_rate": 4.7791171028284434e-05,
+      "loss": 1.3487,
+      "step": 8366
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1771240234375,
+      "learning_rate": 4.779090668781391e-05,
+      "loss": 1.5783,
+      "step": 8367
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9541938304901123,
+      "learning_rate": 4.779064234734338e-05,
+      "loss": 1.4343,
+      "step": 8368
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6581960916519165,
+      "learning_rate": 4.7790378006872855e-05,
+      "loss": 1.4733,
+      "step": 8369
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9849929809570312,
+      "learning_rate": 4.779011366640233e-05,
+      "loss": 1.4321,
+      "step": 8370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6208577156066895,
+      "learning_rate": 4.77898493259318e-05,
+      "loss": 1.3568,
+      "step": 8371
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.843384027481079,
+      "learning_rate": 4.7789584985461275e-05,
+      "loss": 1.482,
+      "step": 8372
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5074721574783325,
+      "learning_rate": 4.778932064499075e-05,
+      "loss": 1.273,
+      "step": 8373
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9038312435150146,
+      "learning_rate": 4.778905630452022e-05,
+      "loss": 1.5486,
+      "step": 8374
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9027255773544312,
+      "learning_rate": 4.77887919640497e-05,
+      "loss": 1.2981,
+      "step": 8375
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5217087268829346,
+      "learning_rate": 4.778852762357917e-05,
+      "loss": 1.3698,
+      "step": 8376
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7686501741409302,
+      "learning_rate": 4.778826328310865e-05,
+      "loss": 1.3275,
+      "step": 8377
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7106071710586548,
+      "learning_rate": 4.778799894263812e-05,
+      "loss": 1.5269,
+      "step": 8378
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4519718885421753,
+      "learning_rate": 4.77877346021676e-05,
+      "loss": 1.2112,
+      "step": 8379
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8169400691986084,
+      "learning_rate": 4.778747026169707e-05,
+      "loss": 1.2605,
+      "step": 8380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.410823941230774,
+      "learning_rate": 4.7787205921226544e-05,
+      "loss": 1.3569,
+      "step": 8381
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9673640727996826,
+      "learning_rate": 4.778694158075602e-05,
+      "loss": 1.4169,
+      "step": 8382
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5068556070327759,
+      "learning_rate": 4.7786677240285485e-05,
+      "loss": 1.3318,
+      "step": 8383
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7520672082901,
+      "learning_rate": 4.7786412899814965e-05,
+      "loss": 1.427,
+      "step": 8384
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9566296339035034,
+      "learning_rate": 4.778614855934444e-05,
+      "loss": 1.4442,
+      "step": 8385
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7213634252548218,
+      "learning_rate": 4.778588421887391e-05,
+      "loss": 1.3178,
+      "step": 8386
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2114782333374023,
+      "learning_rate": 4.7785619878403386e-05,
+      "loss": 1.5963,
+      "step": 8387
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7535631656646729,
+      "learning_rate": 4.778535553793286e-05,
+      "loss": 1.4505,
+      "step": 8388
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0050625801086426,
+      "learning_rate": 4.778509119746233e-05,
+      "loss": 1.3915,
+      "step": 8389
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5159224271774292,
+      "learning_rate": 4.778482685699181e-05,
+      "loss": 1.2006,
+      "step": 8390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6937108039855957,
+      "learning_rate": 4.778456251652128e-05,
+      "loss": 1.2977,
+      "step": 8391
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8329828977584839,
+      "learning_rate": 4.7784298176050754e-05,
+      "loss": 1.5239,
+      "step": 8392
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1298718452453613,
+      "learning_rate": 4.778403383558023e-05,
+      "loss": 1.4143,
+      "step": 8393
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6044392585754395,
+      "learning_rate": 4.77837694951097e-05,
+      "loss": 1.3857,
+      "step": 8394
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8792058229446411,
+      "learning_rate": 4.778350515463918e-05,
+      "loss": 1.3401,
+      "step": 8395
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5751330852508545,
+      "learning_rate": 4.778324081416865e-05,
+      "loss": 1.451,
+      "step": 8396
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4163377285003662,
+      "learning_rate": 4.778297647369813e-05,
+      "loss": 1.9717,
+      "step": 8397
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.077528715133667,
+      "learning_rate": 4.7782712133227595e-05,
+      "loss": 1.1981,
+      "step": 8398
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8986663818359375,
+      "learning_rate": 4.7782447792757075e-05,
+      "loss": 1.4926,
+      "step": 8399
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8446327447891235,
+      "learning_rate": 4.778218345228655e-05,
+      "loss": 1.4011,
+      "step": 8400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5474783182144165,
+      "learning_rate": 4.778191911181602e-05,
+      "loss": 1.3492,
+      "step": 8401
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3860501050949097,
+      "learning_rate": 4.7781654771345496e-05,
+      "loss": 1.339,
+      "step": 8402
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.929563283920288,
+      "learning_rate": 4.778139043087497e-05,
+      "loss": 1.4329,
+      "step": 8403
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.759806752204895,
+      "learning_rate": 4.778112609040444e-05,
+      "loss": 1.2326,
+      "step": 8404
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.564479112625122,
+      "learning_rate": 4.778086174993392e-05,
+      "loss": 1.4247,
+      "step": 8405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6182713508605957,
+      "learning_rate": 4.778059740946339e-05,
+      "loss": 1.36,
+      "step": 8406
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.219984292984009,
+      "learning_rate": 4.7780333068992864e-05,
+      "loss": 1.4382,
+      "step": 8407
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6588809490203857,
+      "learning_rate": 4.778006872852234e-05,
+      "loss": 1.2411,
+      "step": 8408
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9430526494979858,
+      "learning_rate": 4.777980438805181e-05,
+      "loss": 1.3519,
+      "step": 8409
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6104369163513184,
+      "learning_rate": 4.777954004758129e-05,
+      "loss": 1.4572,
+      "step": 8410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.301051616668701,
+      "learning_rate": 4.777927570711076e-05,
+      "loss": 1.5531,
+      "step": 8411
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2343785762786865,
+      "learning_rate": 4.777901136664024e-05,
+      "loss": 1.4798,
+      "step": 8412
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5529707670211792,
+      "learning_rate": 4.7778747026169705e-05,
+      "loss": 1.4117,
+      "step": 8413
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5635724067687988,
+      "learning_rate": 4.7778482685699186e-05,
+      "loss": 1.3254,
+      "step": 8414
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8641867637634277,
+      "learning_rate": 4.777821834522866e-05,
+      "loss": 1.4534,
+      "step": 8415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7195168733596802,
+      "learning_rate": 4.777795400475813e-05,
+      "loss": 1.2198,
+      "step": 8416
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.059126615524292,
+      "learning_rate": 4.7777689664287606e-05,
+      "loss": 1.4484,
+      "step": 8417
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.370356559753418,
+      "learning_rate": 4.777742532381708e-05,
+      "loss": 1.3318,
+      "step": 8418
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5832743644714355,
+      "learning_rate": 4.7777160983346554e-05,
+      "loss": 1.4784,
+      "step": 8419
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4089196920394897,
+      "learning_rate": 4.777689664287603e-05,
+      "loss": 1.3488,
+      "step": 8420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8296297788619995,
+      "learning_rate": 4.77766323024055e-05,
+      "loss": 1.4079,
+      "step": 8421
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6849052906036377,
+      "learning_rate": 4.7776367961934974e-05,
+      "loss": 2.355,
+      "step": 8422
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4913830757141113,
+      "learning_rate": 4.777610362146445e-05,
+      "loss": 1.5924,
+      "step": 8423
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.473742961883545,
+      "learning_rate": 4.777583928099392e-05,
+      "loss": 1.2684,
+      "step": 8424
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8239164352416992,
+      "learning_rate": 4.77755749405234e-05,
+      "loss": 1.3206,
+      "step": 8425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5390568971633911,
+      "learning_rate": 4.777531060005287e-05,
+      "loss": 1.4988,
+      "step": 8426
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8862698078155518,
+      "learning_rate": 4.777504625958235e-05,
+      "loss": 1.5325,
+      "step": 8427
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.048360824584961,
+      "learning_rate": 4.7774781919111816e-05,
+      "loss": 1.4774,
+      "step": 8428
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1043221950531006,
+      "learning_rate": 4.777451757864129e-05,
+      "loss": 1.5584,
+      "step": 8429
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5099202394485474,
+      "learning_rate": 4.777425323817077e-05,
+      "loss": 1.3771,
+      "step": 8430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8067415952682495,
+      "learning_rate": 4.7773988897700236e-05,
+      "loss": 1.4698,
+      "step": 8431
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.619131326675415,
+      "learning_rate": 4.777372455722972e-05,
+      "loss": 1.4918,
+      "step": 8432
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4709293842315674,
+      "learning_rate": 4.7773460216759184e-05,
+      "loss": 1.3997,
+      "step": 8433
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6844581365585327,
+      "learning_rate": 4.7773195876288664e-05,
+      "loss": 1.4776,
+      "step": 8434
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2850040197372437,
+      "learning_rate": 4.777293153581814e-05,
+      "loss": 1.2561,
+      "step": 8435
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4677332639694214,
+      "learning_rate": 4.777266719534761e-05,
+      "loss": 1.6063,
+      "step": 8436
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4393115043640137,
+      "learning_rate": 4.7772402854877085e-05,
+      "loss": 1.1872,
+      "step": 8437
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6552435159683228,
+      "learning_rate": 4.777213851440656e-05,
+      "loss": 1.322,
+      "step": 8438
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5255144834518433,
+      "learning_rate": 4.777187417393603e-05,
+      "loss": 1.3654,
+      "step": 8439
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.37326979637146,
+      "learning_rate": 4.7771609833465505e-05,
+      "loss": 1.5317,
+      "step": 8440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1190972328186035,
+      "learning_rate": 4.777134549299498e-05,
+      "loss": 1.7143,
+      "step": 8441
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.829806804656982,
+      "learning_rate": 4.777108115252445e-05,
+      "loss": 1.7209,
+      "step": 8442
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7885311841964722,
+      "learning_rate": 4.7770816812053926e-05,
+      "loss": 1.4522,
+      "step": 8443
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5140645503997803,
+      "learning_rate": 4.77705524715834e-05,
+      "loss": 1.3492,
+      "step": 8444
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6319007873535156,
+      "learning_rate": 4.777028813111288e-05,
+      "loss": 1.314,
+      "step": 8445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8566958904266357,
+      "learning_rate": 4.777002379064235e-05,
+      "loss": 1.3933,
+      "step": 8446
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2969062328338623,
+      "learning_rate": 4.776975945017183e-05,
+      "loss": 1.4695,
+      "step": 8447
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7640814781188965,
+      "learning_rate": 4.7769495109701294e-05,
+      "loss": 1.3147,
+      "step": 8448
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9109821319580078,
+      "learning_rate": 4.7769230769230774e-05,
+      "loss": 1.3747,
+      "step": 8449
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1373677253723145,
+      "learning_rate": 4.776896642876025e-05,
+      "loss": 1.4637,
+      "step": 8450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.260056495666504,
+      "learning_rate": 4.776870208828972e-05,
+      "loss": 2.7594,
+      "step": 8451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6303960084915161,
+      "learning_rate": 4.7768437747819195e-05,
+      "loss": 1.5099,
+      "step": 8452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7009268999099731,
+      "learning_rate": 4.776817340734867e-05,
+      "loss": 1.327,
+      "step": 8453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.418572425842285,
+      "learning_rate": 4.776790906687814e-05,
+      "loss": 1.4485,
+      "step": 8454
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4898825883865356,
+      "learning_rate": 4.7767644726407616e-05,
+      "loss": 2.1469,
+      "step": 8455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9026637077331543,
+      "learning_rate": 4.776738038593709e-05,
+      "loss": 1.4215,
+      "step": 8456
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6575506925582886,
+      "learning_rate": 4.776711604546656e-05,
+      "loss": 1.2848,
+      "step": 8457
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6632676124572754,
+      "learning_rate": 4.7766851704996036e-05,
+      "loss": 1.4262,
+      "step": 8458
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.373077630996704,
+      "learning_rate": 4.776658736452551e-05,
+      "loss": 1.6057,
+      "step": 8459
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7636457681655884,
+      "learning_rate": 4.776632302405499e-05,
+      "loss": 1.4875,
+      "step": 8460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.328150749206543,
+      "learning_rate": 4.776605868358446e-05,
+      "loss": 1.3817,
+      "step": 8461
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.297027349472046,
+      "learning_rate": 4.776579434311394e-05,
+      "loss": 1.4252,
+      "step": 8462
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4420815706253052,
+      "learning_rate": 4.7765530002643404e-05,
+      "loss": 1.2718,
+      "step": 8463
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.751740574836731,
+      "learning_rate": 4.776526566217288e-05,
+      "loss": 1.4978,
+      "step": 8464
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7586183547973633,
+      "learning_rate": 4.776500132170236e-05,
+      "loss": 1.3644,
+      "step": 8465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.7833638191223145,
+      "learning_rate": 4.7764736981231825e-05,
+      "loss": 1.5276,
+      "step": 8466
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0102052688598633,
+      "learning_rate": 4.7764472640761305e-05,
+      "loss": 1.77,
+      "step": 8467
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.097877264022827,
+      "learning_rate": 4.776420830029077e-05,
+      "loss": 1.3592,
+      "step": 8468
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.424149751663208,
+      "learning_rate": 4.776394395982025e-05,
+      "loss": 1.3655,
+      "step": 8469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9183423519134521,
+      "learning_rate": 4.7763679619349726e-05,
+      "loss": 1.4285,
+      "step": 8470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1547820568084717,
+      "learning_rate": 4.77634152788792e-05,
+      "loss": 1.5839,
+      "step": 8471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.119232654571533,
+      "learning_rate": 4.776315093840867e-05,
+      "loss": 1.4093,
+      "step": 8472
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6863677501678467,
+      "learning_rate": 4.776288659793815e-05,
+      "loss": 1.428,
+      "step": 8473
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8018301725387573,
+      "learning_rate": 4.776262225746762e-05,
+      "loss": 1.5758,
+      "step": 8474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6270349025726318,
+      "learning_rate": 4.7762357916997094e-05,
+      "loss": 1.3468,
+      "step": 8475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7268153429031372,
+      "learning_rate": 4.776209357652657e-05,
+      "loss": 1.4073,
+      "step": 8476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.951058030128479,
+      "learning_rate": 4.776182923605604e-05,
+      "loss": 1.4345,
+      "step": 8477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9563771486282349,
+      "learning_rate": 4.7761564895585515e-05,
+      "loss": 1.4904,
+      "step": 8478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4810471534729004,
+      "learning_rate": 4.776130055511499e-05,
+      "loss": 1.3246,
+      "step": 8479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6206552982330322,
+      "learning_rate": 4.776103621464447e-05,
+      "loss": 1.4214,
+      "step": 8480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5589320659637451,
+      "learning_rate": 4.7760771874173935e-05,
+      "loss": 1.3416,
+      "step": 8481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8163495063781738,
+      "learning_rate": 4.7760507533703416e-05,
+      "loss": 1.561,
+      "step": 8482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.058225393295288,
+      "learning_rate": 4.776024319323288e-05,
+      "loss": 1.538,
+      "step": 8483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5687527656555176,
+      "learning_rate": 4.775997885276236e-05,
+      "loss": 1.493,
+      "step": 8484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1514317989349365,
+      "learning_rate": 4.7759714512291836e-05,
+      "loss": 1.428,
+      "step": 8485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8540575504302979,
+      "learning_rate": 4.775945017182131e-05,
+      "loss": 1.3593,
+      "step": 8486
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1896848678588867,
+      "learning_rate": 4.7759185831350783e-05,
+      "loss": 1.2911,
+      "step": 8487
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0426063537597656,
+      "learning_rate": 4.775892149088026e-05,
+      "loss": 1.5484,
+      "step": 8488
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5374425649642944,
+      "learning_rate": 4.775865715040973e-05,
+      "loss": 1.3431,
+      "step": 8489
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.091796875,
+      "learning_rate": 4.7758392809939204e-05,
+      "loss": 1.3232,
+      "step": 8490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.859734296798706,
+      "learning_rate": 4.775812846946868e-05,
+      "loss": 1.2851,
+      "step": 8491
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.737462043762207,
+      "learning_rate": 4.775786412899815e-05,
+      "loss": 2.4241,
+      "step": 8492
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.848176121711731,
+      "learning_rate": 4.7757599788527625e-05,
+      "loss": 1.4553,
+      "step": 8493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8728655576705933,
+      "learning_rate": 4.77573354480571e-05,
+      "loss": 1.5562,
+      "step": 8494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9957512617111206,
+      "learning_rate": 4.775707110758658e-05,
+      "loss": 1.3834,
+      "step": 8495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4220339059829712,
+      "learning_rate": 4.7756806767116046e-05,
+      "loss": 2.0075,
+      "step": 8496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0616798400878906,
+      "learning_rate": 4.7756542426645526e-05,
+      "loss": 1.4905,
+      "step": 8497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3814077377319336,
+      "learning_rate": 4.775627808617499e-05,
+      "loss": 1.3506,
+      "step": 8498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4639906883239746,
+      "learning_rate": 4.775601374570447e-05,
+      "loss": 1.1733,
+      "step": 8499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5455049276351929,
+      "learning_rate": 4.775574940523395e-05,
+      "loss": 1.3204,
+      "step": 8500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6093316078186035,
+      "learning_rate": 4.7755485064763413e-05,
+      "loss": 1.5288,
+      "step": 8501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.247256278991699,
+      "learning_rate": 4.7755220724292894e-05,
+      "loss": 1.4326,
+      "step": 8502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7475184202194214,
+      "learning_rate": 4.775495638382236e-05,
+      "loss": 1.3196,
+      "step": 8503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8701509237289429,
+      "learning_rate": 4.775469204335184e-05,
+      "loss": 1.4424,
+      "step": 8504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4840946197509766,
+      "learning_rate": 4.7754427702881315e-05,
+      "loss": 1.318,
+      "step": 8505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7931088209152222,
+      "learning_rate": 4.775416336241079e-05,
+      "loss": 1.3957,
+      "step": 8506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.057534694671631,
+      "learning_rate": 4.775389902194026e-05,
+      "loss": 1.4332,
+      "step": 8507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5310745239257812,
+      "learning_rate": 4.7753634681469735e-05,
+      "loss": 1.3742,
+      "step": 8508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.920225739479065,
+      "learning_rate": 4.775337034099921e-05,
+      "loss": 1.3842,
+      "step": 8509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7083951234817505,
+      "learning_rate": 4.775310600052868e-05,
+      "loss": 1.3955,
+      "step": 8510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.065639019012451,
+      "learning_rate": 4.7752841660058156e-05,
+      "loss": 1.4965,
+      "step": 8511
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.17134428024292,
+      "learning_rate": 4.775257731958763e-05,
+      "loss": 1.4845,
+      "step": 8512
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.548410177230835,
+      "learning_rate": 4.77523129791171e-05,
+      "loss": 1.4305,
+      "step": 8513
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8961461782455444,
+      "learning_rate": 4.775204863864658e-05,
+      "loss": 1.4822,
+      "step": 8514
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.813829779624939,
+      "learning_rate": 4.775178429817606e-05,
+      "loss": 1.3773,
+      "step": 8515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.026472806930542,
+      "learning_rate": 4.7751519957705524e-05,
+      "loss": 1.3907,
+      "step": 8516
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6190162897109985,
+      "learning_rate": 4.7751255617235004e-05,
+      "loss": 1.3877,
+      "step": 8517
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.068399429321289,
+      "learning_rate": 4.775099127676447e-05,
+      "loss": 1.4626,
+      "step": 8518
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7386622428894043,
+      "learning_rate": 4.775072693629395e-05,
+      "loss": 1.3681,
+      "step": 8519
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.591489315032959,
+      "learning_rate": 4.7750462595823425e-05,
+      "loss": 1.3268,
+      "step": 8520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.141873359680176,
+      "learning_rate": 4.77501982553529e-05,
+      "loss": 1.4525,
+      "step": 8521
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.651975393295288,
+      "learning_rate": 4.774993391488237e-05,
+      "loss": 1.3417,
+      "step": 8522
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.049616575241089,
+      "learning_rate": 4.7749669574411846e-05,
+      "loss": 1.3842,
+      "step": 8523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6224886178970337,
+      "learning_rate": 4.774940523394132e-05,
+      "loss": 1.317,
+      "step": 8524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9622199535369873,
+      "learning_rate": 4.774914089347079e-05,
+      "loss": 1.4645,
+      "step": 8525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9653286933898926,
+      "learning_rate": 4.7748876553000266e-05,
+      "loss": 1.3751,
+      "step": 8526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.756283164024353,
+      "learning_rate": 4.774861221252974e-05,
+      "loss": 1.3782,
+      "step": 8527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.780847191810608,
+      "learning_rate": 4.7748347872059213e-05,
+      "loss": 1.4171,
+      "step": 8528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.127122640609741,
+      "learning_rate": 4.774808353158869e-05,
+      "loss": 1.391,
+      "step": 8529
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.587888479232788,
+      "learning_rate": 4.774781919111817e-05,
+      "loss": 1.3171,
+      "step": 8530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5362863540649414,
+      "learning_rate": 4.7747554850647634e-05,
+      "loss": 1.3265,
+      "step": 8531
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8233709335327148,
+      "learning_rate": 4.7747290510177114e-05,
+      "loss": 1.2802,
+      "step": 8532
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.724384069442749,
+      "learning_rate": 4.774702616970658e-05,
+      "loss": 1.371,
+      "step": 8533
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9820339679718018,
+      "learning_rate": 4.774676182923606e-05,
+      "loss": 1.3359,
+      "step": 8534
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7671434879302979,
+      "learning_rate": 4.7746497488765535e-05,
+      "loss": 1.3826,
+      "step": 8535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8877044916152954,
+      "learning_rate": 4.7746233148295e-05,
+      "loss": 1.4212,
+      "step": 8536
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6429840326309204,
+      "learning_rate": 4.774596880782448e-05,
+      "loss": 1.2515,
+      "step": 8537
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9920988082885742,
+      "learning_rate": 4.774570446735395e-05,
+      "loss": 1.4702,
+      "step": 8538
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0630156993865967,
+      "learning_rate": 4.774544012688343e-05,
+      "loss": 1.3827,
+      "step": 8539
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5286803245544434,
+      "learning_rate": 4.77451757864129e-05,
+      "loss": 1.1555,
+      "step": 8540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.615519404411316,
+      "learning_rate": 4.774491144594238e-05,
+      "loss": 1.3148,
+      "step": 8541
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4771125316619873,
+      "learning_rate": 4.774464710547185e-05,
+      "loss": 1.473,
+      "step": 8542
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.325908899307251,
+      "learning_rate": 4.7744382765001324e-05,
+      "loss": 1.4217,
+      "step": 8543
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6734358072280884,
+      "learning_rate": 4.77441184245308e-05,
+      "loss": 1.5002,
+      "step": 8544
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.389521360397339,
+      "learning_rate": 4.774385408406028e-05,
+      "loss": 1.5616,
+      "step": 8545
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.831109881401062,
+      "learning_rate": 4.7743589743589744e-05,
+      "loss": 1.5041,
+      "step": 8546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.025660991668701,
+      "learning_rate": 4.774332540311922e-05,
+      "loss": 1.4242,
+      "step": 8547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6666538715362549,
+      "learning_rate": 4.774306106264869e-05,
+      "loss": 1.3845,
+      "step": 8548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8368525505065918,
+      "learning_rate": 4.7742796722178165e-05,
+      "loss": 1.2674,
+      "step": 8549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.343019962310791,
+      "learning_rate": 4.7742532381707646e-05,
+      "loss": 1.527,
+      "step": 8550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4630554914474487,
+      "learning_rate": 4.774226804123711e-05,
+      "loss": 1.9193,
+      "step": 8551
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.302304983139038,
+      "learning_rate": 4.774200370076659e-05,
+      "loss": 1.5005,
+      "step": 8552
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8351490497589111,
+      "learning_rate": 4.774173936029606e-05,
+      "loss": 1.5613,
+      "step": 8553
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.150634527206421,
+      "learning_rate": 4.774147501982554e-05,
+      "loss": 1.3874,
+      "step": 8554
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.018548011779785,
+      "learning_rate": 4.7741210679355013e-05,
+      "loss": 1.4631,
+      "step": 8555
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0677378177642822,
+      "learning_rate": 4.774094633888449e-05,
+      "loss": 1.3174,
+      "step": 8556
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9459774494171143,
+      "learning_rate": 4.774068199841396e-05,
+      "loss": 1.4642,
+      "step": 8557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1573665142059326,
+      "learning_rate": 4.7740417657943434e-05,
+      "loss": 1.4803,
+      "step": 8558
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5278096199035645,
+      "learning_rate": 4.774015331747291e-05,
+      "loss": 1.4118,
+      "step": 8559
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8153343200683594,
+      "learning_rate": 4.773988897700238e-05,
+      "loss": 1.4666,
+      "step": 8560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.003507375717163,
+      "learning_rate": 4.7739624636531855e-05,
+      "loss": 1.5781,
+      "step": 8561
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6568461656570435,
+      "learning_rate": 4.773936029606133e-05,
+      "loss": 1.3658,
+      "step": 8562
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6068572998046875,
+      "learning_rate": 4.77390959555908e-05,
+      "loss": 1.4584,
+      "step": 8563
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7118079662322998,
+      "learning_rate": 4.7738831615120276e-05,
+      "loss": 1.3182,
+      "step": 8564
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7038938999176025,
+      "learning_rate": 4.7738567274649756e-05,
+      "loss": 1.4866,
+      "step": 8565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8129291534423828,
+      "learning_rate": 4.773830293417922e-05,
+      "loss": 1.5364,
+      "step": 8566
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9018023014068604,
+      "learning_rate": 4.77380385937087e-05,
+      "loss": 1.3887,
+      "step": 8567
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6865167617797852,
+      "learning_rate": 4.773777425323817e-05,
+      "loss": 1.3634,
+      "step": 8568
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7502326965332031,
+      "learning_rate": 4.773750991276765e-05,
+      "loss": 2.026,
+      "step": 8569
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.978610873222351,
+      "learning_rate": 4.7737245572297124e-05,
+      "loss": 1.4292,
+      "step": 8570
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9288057088851929,
+      "learning_rate": 4.77369812318266e-05,
+      "loss": 1.3501,
+      "step": 8571
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.219231367111206,
+      "learning_rate": 4.773671689135607e-05,
+      "loss": 1.5479,
+      "step": 8572
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7449499368667603,
+      "learning_rate": 4.773645255088554e-05,
+      "loss": 1.4868,
+      "step": 8573
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.456263780593872,
+      "learning_rate": 4.773618821041502e-05,
+      "loss": 1.2982,
+      "step": 8574
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8473769426345825,
+      "learning_rate": 4.773592386994449e-05,
+      "loss": 1.3975,
+      "step": 8575
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.059465169906616,
+      "learning_rate": 4.7735659529473965e-05,
+      "loss": 1.3981,
+      "step": 8576
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2866734266281128,
+      "learning_rate": 4.773539518900344e-05,
+      "loss": 2.1192,
+      "step": 8577
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6440802812576294,
+      "learning_rate": 4.773513084853291e-05,
+      "loss": 1.2426,
+      "step": 8578
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6466500759124756,
+      "learning_rate": 4.7734866508062386e-05,
+      "loss": 2.2272,
+      "step": 8579
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8029791116714478,
+      "learning_rate": 4.7734602167591866e-05,
+      "loss": 1.3982,
+      "step": 8580
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7820053100585938,
+      "learning_rate": 4.773433782712133e-05,
+      "loss": 1.444,
+      "step": 8581
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4995321035385132,
+      "learning_rate": 4.7734073486650807e-05,
+      "loss": 1.4027,
+      "step": 8582
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8156757354736328,
+      "learning_rate": 4.773380914618028e-05,
+      "loss": 1.6431,
+      "step": 8583
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9216279983520508,
+      "learning_rate": 4.7733544805709754e-05,
+      "loss": 1.3104,
+      "step": 8584
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.650805950164795,
+      "learning_rate": 4.7733280465239234e-05,
+      "loss": 1.3867,
+      "step": 8585
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8753902912139893,
+      "learning_rate": 4.77330161247687e-05,
+      "loss": 1.4234,
+      "step": 8586
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7090024948120117,
+      "learning_rate": 4.773275178429818e-05,
+      "loss": 1.231,
+      "step": 8587
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6158668994903564,
+      "learning_rate": 4.773248744382765e-05,
+      "loss": 1.4349,
+      "step": 8588
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5632296800613403,
+      "learning_rate": 4.773222310335713e-05,
+      "loss": 1.3186,
+      "step": 8589
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4906212091445923,
+      "learning_rate": 4.77319587628866e-05,
+      "loss": 1.1517,
+      "step": 8590
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.734694480895996,
+      "learning_rate": 4.7731694422416076e-05,
+      "loss": 1.4699,
+      "step": 8591
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.369840145111084,
+      "learning_rate": 4.773143008194555e-05,
+      "loss": 1.3509,
+      "step": 8592
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5825780630111694,
+      "learning_rate": 4.773116574147502e-05,
+      "loss": 1.4514,
+      "step": 8593
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4381916522979736,
+      "learning_rate": 4.7730901401004496e-05,
+      "loss": 1.6081,
+      "step": 8594
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.698498249053955,
+      "learning_rate": 4.773063706053397e-05,
+      "loss": 1.5459,
+      "step": 8595
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8351300954818726,
+      "learning_rate": 4.773037272006344e-05,
+      "loss": 1.5141,
+      "step": 8596
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9488400220870972,
+      "learning_rate": 4.773010837959292e-05,
+      "loss": 1.476,
+      "step": 8597
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5306442975997925,
+      "learning_rate": 4.772984403912239e-05,
+      "loss": 1.3285,
+      "step": 8598
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6883198022842407,
+      "learning_rate": 4.7729579698651864e-05,
+      "loss": 1.6374,
+      "step": 8599
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.381324291229248,
+      "learning_rate": 4.7729315358181344e-05,
+      "loss": 1.4286,
+      "step": 8600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7268990278244019,
+      "learning_rate": 4.772905101771081e-05,
+      "loss": 1.4927,
+      "step": 8601
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9519095420837402,
+      "learning_rate": 4.772878667724029e-05,
+      "loss": 1.3809,
+      "step": 8602
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5016953945159912,
+      "learning_rate": 4.772852233676976e-05,
+      "loss": 1.4765,
+      "step": 8603
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3685991764068604,
+      "learning_rate": 4.772825799629924e-05,
+      "loss": 1.4987,
+      "step": 8604
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9244827032089233,
+      "learning_rate": 4.772799365582871e-05,
+      "loss": 1.4946,
+      "step": 8605
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7976433038711548,
+      "learning_rate": 4.7727729315358186e-05,
+      "loss": 1.5599,
+      "step": 8606
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2701480388641357,
+      "learning_rate": 4.772746497488766e-05,
+      "loss": 1.4412,
+      "step": 8607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.784895658493042,
+      "learning_rate": 4.772720063441713e-05,
+      "loss": 1.5481,
+      "step": 8608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5578367710113525,
+      "learning_rate": 4.7726936293946607e-05,
+      "loss": 1.4077,
+      "step": 8609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7594302892684937,
+      "learning_rate": 4.772667195347608e-05,
+      "loss": 1.4302,
+      "step": 8610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.265252113342285,
+      "learning_rate": 4.7726407613005554e-05,
+      "loss": 1.6447,
+      "step": 8611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.598236560821533,
+      "learning_rate": 4.772614327253503e-05,
+      "loss": 1.3647,
+      "step": 8612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.672574520111084,
+      "learning_rate": 4.77258789320645e-05,
+      "loss": 1.3625,
+      "step": 8613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.21756649017334,
+      "learning_rate": 4.7725614591593974e-05,
+      "loss": 1.4912,
+      "step": 8614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.789411187171936,
+      "learning_rate": 4.7725350251123455e-05,
+      "loss": 1.3182,
+      "step": 8615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7805672883987427,
+      "learning_rate": 4.772508591065292e-05,
+      "loss": 1.5097,
+      "step": 8616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.772804856300354,
+      "learning_rate": 4.77248215701824e-05,
+      "loss": 1.3213,
+      "step": 8617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8255934715270996,
+      "learning_rate": 4.772455722971187e-05,
+      "loss": 1.417,
+      "step": 8618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0925443172454834,
+      "learning_rate": 4.772429288924134e-05,
+      "loss": 1.4818,
+      "step": 8619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.399364709854126,
+      "learning_rate": 4.772402854877082e-05,
+      "loss": 1.4125,
+      "step": 8620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7890353202819824,
+      "learning_rate": 4.772376420830029e-05,
+      "loss": 1.4521,
+      "step": 8621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5670008659362793,
+      "learning_rate": 4.772349986782977e-05,
+      "loss": 1.3785,
+      "step": 8622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4909836053848267,
+      "learning_rate": 4.7723235527359237e-05,
+      "loss": 1.4268,
+      "step": 8623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.091538190841675,
+      "learning_rate": 4.772297118688872e-05,
+      "loss": 1.4323,
+      "step": 8624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5458043813705444,
+      "learning_rate": 4.772270684641819e-05,
+      "loss": 1.4458,
+      "step": 8625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4993782043457031,
+      "learning_rate": 4.7722442505947664e-05,
+      "loss": 1.41,
+      "step": 8626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7375521659851074,
+      "learning_rate": 4.772217816547714e-05,
+      "loss": 1.3999,
+      "step": 8627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6506996154785156,
+      "learning_rate": 4.772191382500661e-05,
+      "loss": 1.5178,
+      "step": 8628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.698262095451355,
+      "learning_rate": 4.7721649484536085e-05,
+      "loss": 1.2728,
+      "step": 8629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7875055074691772,
+      "learning_rate": 4.772138514406556e-05,
+      "loss": 1.2151,
+      "step": 8630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8956786394119263,
+      "learning_rate": 4.772112080359503e-05,
+      "loss": 1.3773,
+      "step": 8631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6732250452041626,
+      "learning_rate": 4.7720856463124505e-05,
+      "loss": 1.4041,
+      "step": 8632
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3805735111236572,
+      "learning_rate": 4.772059212265398e-05,
+      "loss": 1.2525,
+      "step": 8633
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.640608310699463,
+      "learning_rate": 4.772032778218345e-05,
+      "loss": 1.2496,
+      "step": 8634
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5390517711639404,
+      "learning_rate": 4.772006344171293e-05,
+      "loss": 1.3004,
+      "step": 8635
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1809651851654053,
+      "learning_rate": 4.77197991012424e-05,
+      "loss": 1.3653,
+      "step": 8636
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.019861936569214,
+      "learning_rate": 4.771953476077188e-05,
+      "loss": 1.4985,
+      "step": 8637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.51957106590271,
+      "learning_rate": 4.771927042030135e-05,
+      "loss": 1.6622,
+      "step": 8638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8437442779541016,
+      "learning_rate": 4.771900607983083e-05,
+      "loss": 1.3709,
+      "step": 8639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.712791085243225,
+      "learning_rate": 4.77187417393603e-05,
+      "loss": 1.2694,
+      "step": 8640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9530540704727173,
+      "learning_rate": 4.7718477398889774e-05,
+      "loss": 1.6186,
+      "step": 8641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.863860011100769,
+      "learning_rate": 4.771821305841925e-05,
+      "loss": 1.2666,
+      "step": 8642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1580379009246826,
+      "learning_rate": 4.771794871794872e-05,
+      "loss": 1.5156,
+      "step": 8643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7668930292129517,
+      "learning_rate": 4.7717684377478195e-05,
+      "loss": 1.3122,
+      "step": 8644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9672787189483643,
+      "learning_rate": 4.771742003700767e-05,
+      "loss": 1.5912,
+      "step": 8645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9556940793991089,
+      "learning_rate": 4.771715569653714e-05,
+      "loss": 1.5311,
+      "step": 8646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6081843376159668,
+      "learning_rate": 4.7716891356066616e-05,
+      "loss": 1.2728,
+      "step": 8647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8273512125015259,
+      "learning_rate": 4.771662701559609e-05,
+      "loss": 1.4487,
+      "step": 8648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8081766366958618,
+      "learning_rate": 4.771636267512556e-05,
+      "loss": 1.3808,
+      "step": 8649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9309170246124268,
+      "learning_rate": 4.771609833465504e-05,
+      "loss": 1.495,
+      "step": 8650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5539155006408691,
+      "learning_rate": 4.771583399418451e-05,
+      "loss": 1.3349,
+      "step": 8651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.719404697418213,
+      "learning_rate": 4.771556965371399e-05,
+      "loss": 1.4746,
+      "step": 8652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.546717405319214,
+      "learning_rate": 4.771530531324346e-05,
+      "loss": 1.6077,
+      "step": 8653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6314597129821777,
+      "learning_rate": 4.771504097277293e-05,
+      "loss": 1.3838,
+      "step": 8654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.679882287979126,
+      "learning_rate": 4.771477663230241e-05,
+      "loss": 1.457,
+      "step": 8655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6158453226089478,
+      "learning_rate": 4.771451229183188e-05,
+      "loss": 1.4774,
+      "step": 8656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8915106058120728,
+      "learning_rate": 4.771424795136136e-05,
+      "loss": 1.4127,
+      "step": 8657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6921844482421875,
+      "learning_rate": 4.7713983610890825e-05,
+      "loss": 1.2931,
+      "step": 8658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7261117696762085,
+      "learning_rate": 4.7713719270420305e-05,
+      "loss": 1.3475,
+      "step": 8659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.465744137763977,
+      "learning_rate": 4.771345492994978e-05,
+      "loss": 1.1839,
+      "step": 8660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9329191446304321,
+      "learning_rate": 4.771319058947925e-05,
+      "loss": 1.4589,
+      "step": 8661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3337154388427734,
+      "learning_rate": 4.7712926249008726e-05,
+      "loss": 1.4653,
+      "step": 8662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7373899221420288,
+      "learning_rate": 4.77126619085382e-05,
+      "loss": 1.4455,
+      "step": 8663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.05975604057312,
+      "learning_rate": 4.771239756806767e-05,
+      "loss": 1.4116,
+      "step": 8664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.915482997894287,
+      "learning_rate": 4.771213322759715e-05,
+      "loss": 1.462,
+      "step": 8665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6953800916671753,
+      "learning_rate": 4.771186888712662e-05,
+      "loss": 1.4838,
+      "step": 8666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5169626474380493,
+      "learning_rate": 4.7711604546656094e-05,
+      "loss": 1.3556,
+      "step": 8667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.591649055480957,
+      "learning_rate": 4.771134020618557e-05,
+      "loss": 1.286,
+      "step": 8668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7169744968414307,
+      "learning_rate": 4.771107586571504e-05,
+      "loss": 1.3443,
+      "step": 8669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.071589708328247,
+      "learning_rate": 4.771081152524452e-05,
+      "loss": 1.4968,
+      "step": 8670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.119358777999878,
+      "learning_rate": 4.771054718477399e-05,
+      "loss": 1.5067,
+      "step": 8671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7451212406158447,
+      "learning_rate": 4.771028284430347e-05,
+      "loss": 1.4666,
+      "step": 8672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0829198360443115,
+      "learning_rate": 4.7710018503832935e-05,
+      "loss": 1.4243,
+      "step": 8673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7344701290130615,
+      "learning_rate": 4.7709754163362416e-05,
+      "loss": 1.4852,
+      "step": 8674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5441453456878662,
+      "learning_rate": 4.770948982289189e-05,
+      "loss": 1.2943,
+      "step": 8675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0067458152770996,
+      "learning_rate": 4.770922548242136e-05,
+      "loss": 1.4195,
+      "step": 8676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.013136863708496,
+      "learning_rate": 4.7708961141950836e-05,
+      "loss": 1.5608,
+      "step": 8677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.968267560005188,
+      "learning_rate": 4.770869680148031e-05,
+      "loss": 1.358,
+      "step": 8678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6206883192062378,
+      "learning_rate": 4.7708432461009784e-05,
+      "loss": 1.348,
+      "step": 8679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4727321863174438,
+      "learning_rate": 4.770816812053926e-05,
+      "loss": 1.2029,
+      "step": 8680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.722499966621399,
+      "learning_rate": 4.770790378006873e-05,
+      "loss": 1.498,
+      "step": 8681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.644144892692566,
+      "learning_rate": 4.7707639439598204e-05,
+      "loss": 1.4105,
+      "step": 8682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.232048511505127,
+      "learning_rate": 4.770737509912768e-05,
+      "loss": 1.4691,
+      "step": 8683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.227029323577881,
+      "learning_rate": 4.770711075865715e-05,
+      "loss": 1.1847,
+      "step": 8684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7848360538482666,
+      "learning_rate": 4.770684641818663e-05,
+      "loss": 1.5117,
+      "step": 8685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5657330751419067,
+      "learning_rate": 4.77065820777161e-05,
+      "loss": 1.3929,
+      "step": 8686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7193056344985962,
+      "learning_rate": 4.770631773724558e-05,
+      "loss": 1.3195,
+      "step": 8687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7362737655639648,
+      "learning_rate": 4.7706053396775046e-05,
+      "loss": 1.5304,
+      "step": 8688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.416797161102295,
+      "learning_rate": 4.7705789056304526e-05,
+      "loss": 1.2969,
+      "step": 8689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.509336233139038,
+      "learning_rate": 4.7705524715834e-05,
+      "loss": 1.4955,
+      "step": 8690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6457639932632446,
+      "learning_rate": 4.7705260375363466e-05,
+      "loss": 1.3157,
+      "step": 8691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6202783584594727,
+      "learning_rate": 4.770499603489295e-05,
+      "loss": 1.5618,
+      "step": 8692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6096305847167969,
+      "learning_rate": 4.7704731694422414e-05,
+      "loss": 1.3742,
+      "step": 8693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.035011053085327,
+      "learning_rate": 4.7704467353951894e-05,
+      "loss": 1.3068,
+      "step": 8694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.352546453475952,
+      "learning_rate": 4.770420301348137e-05,
+      "loss": 1.4229,
+      "step": 8695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9593534469604492,
+      "learning_rate": 4.770393867301084e-05,
+      "loss": 1.5508,
+      "step": 8696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8100559711456299,
+      "learning_rate": 4.7703674332540315e-05,
+      "loss": 1.3886,
+      "step": 8697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.833411693572998,
+      "learning_rate": 4.770340999206979e-05,
+      "loss": 1.4544,
+      "step": 8698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3809068202972412,
+      "learning_rate": 4.770314565159926e-05,
+      "loss": 2.3986,
+      "step": 8699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2948458194732666,
+      "learning_rate": 4.7702881311128735e-05,
+      "loss": 1.3951,
+      "step": 8700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4743930101394653,
+      "learning_rate": 4.770261697065821e-05,
+      "loss": 1.3828,
+      "step": 8701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.918447732925415,
+      "learning_rate": 4.770235263018768e-05,
+      "loss": 1.5107,
+      "step": 8702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.849342703819275,
+      "learning_rate": 4.7702088289717156e-05,
+      "loss": 1.8066,
+      "step": 8703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0303893089294434,
+      "learning_rate": 4.770182394924663e-05,
+      "loss": 1.4184,
+      "step": 8704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.970592737197876,
+      "learning_rate": 4.770155960877611e-05,
+      "loss": 1.2674,
+      "step": 8705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8643898963928223,
+      "learning_rate": 4.770129526830558e-05,
+      "loss": 1.5015,
+      "step": 8706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.256549596786499,
+      "learning_rate": 4.770103092783506e-05,
+      "loss": 1.4459,
+      "step": 8707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.879740834236145,
+      "learning_rate": 4.7700766587364524e-05,
+      "loss": 1.4591,
+      "step": 8708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8660929203033447,
+      "learning_rate": 4.7700502246894004e-05,
+      "loss": 1.6005,
+      "step": 8709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.154355764389038,
+      "learning_rate": 4.770023790642348e-05,
+      "loss": 1.4688,
+      "step": 8710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5217022895812988,
+      "learning_rate": 4.769997356595295e-05,
+      "loss": 1.2084,
+      "step": 8711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5190184116363525,
+      "learning_rate": 4.7699709225482425e-05,
+      "loss": 1.3049,
+      "step": 8712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0676450729370117,
+      "learning_rate": 4.76994448850119e-05,
+      "loss": 1.5254,
+      "step": 8713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1707918643951416,
+      "learning_rate": 4.769918054454137e-05,
+      "loss": 1.3952,
+      "step": 8714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3060319423675537,
+      "learning_rate": 4.7698916204070846e-05,
+      "loss": 1.3325,
+      "step": 8715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.149838447570801,
+      "learning_rate": 4.769865186360032e-05,
+      "loss": 1.565,
+      "step": 8716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2504611015319824,
+      "learning_rate": 4.769838752312979e-05,
+      "loss": 1.4157,
+      "step": 8717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7664661407470703,
+      "learning_rate": 4.7698123182659266e-05,
+      "loss": 1.5575,
+      "step": 8718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.141925573348999,
+      "learning_rate": 4.769785884218874e-05,
+      "loss": 1.5865,
+      "step": 8719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7526228427886963,
+      "learning_rate": 4.769759450171822e-05,
+      "loss": 1.5605,
+      "step": 8720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9733132123947144,
+      "learning_rate": 4.769733016124769e-05,
+      "loss": 1.3785,
+      "step": 8721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5573920011520386,
+      "learning_rate": 4.769706582077717e-05,
+      "loss": 1.3056,
+      "step": 8722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9439314603805542,
+      "learning_rate": 4.7696801480306634e-05,
+      "loss": 1.4489,
+      "step": 8723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9626044034957886,
+      "learning_rate": 4.7696537139836115e-05,
+      "loss": 1.3363,
+      "step": 8724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8482258319854736,
+      "learning_rate": 4.769627279936559e-05,
+      "loss": 1.3003,
+      "step": 8725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.702743411064148,
+      "learning_rate": 4.769600845889506e-05,
+      "loss": 1.3534,
+      "step": 8726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7990301847457886,
+      "learning_rate": 4.7695744118424535e-05,
+      "loss": 1.3166,
+      "step": 8727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.642332911491394,
+      "learning_rate": 4.7695479777954e-05,
+      "loss": 1.344,
+      "step": 8728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8166968822479248,
+      "learning_rate": 4.769521543748348e-05,
+      "loss": 1.5337,
+      "step": 8729
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4959710836410522,
+      "learning_rate": 4.7694951097012956e-05,
+      "loss": 1.4153,
+      "step": 8730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9886113405227661,
+      "learning_rate": 4.769468675654243e-05,
+      "loss": 1.4475,
+      "step": 8731
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6286896467208862,
+      "learning_rate": 4.76944224160719e-05,
+      "loss": 1.3711,
+      "step": 8732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8112374544143677,
+      "learning_rate": 4.769415807560138e-05,
+      "loss": 1.3687,
+      "step": 8733
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6906036138534546,
+      "learning_rate": 4.769389373513085e-05,
+      "loss": 1.3051,
+      "step": 8734
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4932128190994263,
+      "learning_rate": 4.769362939466033e-05,
+      "loss": 1.4672,
+      "step": 8735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.852671504020691,
+      "learning_rate": 4.76933650541898e-05,
+      "loss": 1.3537,
+      "step": 8736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4247735738754272,
+      "learning_rate": 4.769310071371927e-05,
+      "loss": 1.2447,
+      "step": 8737
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8086788654327393,
+      "learning_rate": 4.7692836373248745e-05,
+      "loss": 1.2943,
+      "step": 8738
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1673710346221924,
+      "learning_rate": 4.769257203277822e-05,
+      "loss": 1.6013,
+      "step": 8739
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.729941725730896,
+      "learning_rate": 4.76923076923077e-05,
+      "loss": 1.499,
+      "step": 8740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6432446241378784,
+      "learning_rate": 4.7692043351837165e-05,
+      "loss": 1.329,
+      "step": 8741
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.446427583694458,
+      "learning_rate": 4.7691779011366646e-05,
+      "loss": 1.2535,
+      "step": 8742
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0310018062591553,
+      "learning_rate": 4.769151467089611e-05,
+      "loss": 1.4453,
+      "step": 8743
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6491094827651978,
+      "learning_rate": 4.769125033042559e-05,
+      "loss": 1.4103,
+      "step": 8744
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6695209741592407,
+      "learning_rate": 4.7690985989955066e-05,
+      "loss": 1.349,
+      "step": 8745
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.006754159927368,
+      "learning_rate": 4.769072164948454e-05,
+      "loss": 2.008,
+      "step": 8746
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2961947917938232,
+      "learning_rate": 4.7690457309014014e-05,
+      "loss": 1.3953,
+      "step": 8747
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6767058372497559,
+      "learning_rate": 4.769019296854349e-05,
+      "loss": 1.3592,
+      "step": 8748
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8946369886398315,
+      "learning_rate": 4.768992862807296e-05,
+      "loss": 1.3992,
+      "step": 8749
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.272351026535034,
+      "learning_rate": 4.7689664287602434e-05,
+      "loss": 1.5814,
+      "step": 8750
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0494940280914307,
+      "learning_rate": 4.768939994713191e-05,
+      "loss": 1.4967,
+      "step": 8751
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3594446182250977,
+      "learning_rate": 4.768913560666138e-05,
+      "loss": 1.5156,
+      "step": 8752
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8606420755386353,
+      "learning_rate": 4.7688871266190855e-05,
+      "loss": 1.5705,
+      "step": 8753
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0295424461364746,
+      "learning_rate": 4.768860692572033e-05,
+      "loss": 1.4511,
+      "step": 8754
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6346242427825928,
+      "learning_rate": 4.768834258524981e-05,
+      "loss": 1.3169,
+      "step": 8755
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3630026578903198,
+      "learning_rate": 4.7688078244779276e-05,
+      "loss": 1.2768,
+      "step": 8756
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.224740743637085,
+      "learning_rate": 4.7687813904308756e-05,
+      "loss": 1.4872,
+      "step": 8757
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1133384704589844,
+      "learning_rate": 4.768754956383822e-05,
+      "loss": 1.5588,
+      "step": 8758
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6416934728622437,
+      "learning_rate": 4.76872852233677e-05,
+      "loss": 1.36,
+      "step": 8759
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3946001529693604,
+      "learning_rate": 4.768702088289718e-05,
+      "loss": 1.6865,
+      "step": 8760
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7547663450241089,
+      "learning_rate": 4.768675654242665e-05,
+      "loss": 1.4913,
+      "step": 8761
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6279798746109009,
+      "learning_rate": 4.7686492201956124e-05,
+      "loss": 1.3244,
+      "step": 8762
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8831660747528076,
+      "learning_rate": 4.768622786148559e-05,
+      "loss": 1.4715,
+      "step": 8763
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9170427322387695,
+      "learning_rate": 4.768596352101507e-05,
+      "loss": 1.3521,
+      "step": 8764
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9834542274475098,
+      "learning_rate": 4.7685699180544545e-05,
+      "loss": 1.4551,
+      "step": 8765
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.027599811553955,
+      "learning_rate": 4.768543484007402e-05,
+      "loss": 1.5456,
+      "step": 8766
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.529817819595337,
+      "learning_rate": 4.768517049960349e-05,
+      "loss": 1.3142,
+      "step": 8767
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.96768057346344,
+      "learning_rate": 4.7684906159132965e-05,
+      "loss": 1.4506,
+      "step": 8768
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9895919561386108,
+      "learning_rate": 4.768464181866244e-05,
+      "loss": 1.2994,
+      "step": 8769
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.503353476524353,
+      "learning_rate": 4.768437747819192e-05,
+      "loss": 1.2986,
+      "step": 8770
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8539212942123413,
+      "learning_rate": 4.7684113137721386e-05,
+      "loss": 1.3412,
+      "step": 8771
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8016849756240845,
+      "learning_rate": 4.768384879725086e-05,
+      "loss": 1.4734,
+      "step": 8772
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5569915771484375,
+      "learning_rate": 4.768358445678033e-05,
+      "loss": 1.3384,
+      "step": 8773
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9853227138519287,
+      "learning_rate": 4.768332011630981e-05,
+      "loss": 1.8036,
+      "step": 8774
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.632283091545105,
+      "learning_rate": 4.768305577583929e-05,
+      "loss": 1.4875,
+      "step": 8775
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.792608618736267,
+      "learning_rate": 4.7682791435368754e-05,
+      "loss": 1.4242,
+      "step": 8776
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9395729303359985,
+      "learning_rate": 4.7682527094898234e-05,
+      "loss": 1.3539,
+      "step": 8777
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.576677680015564,
+      "learning_rate": 4.76822627544277e-05,
+      "loss": 2.4041,
+      "step": 8778
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4438962936401367,
+      "learning_rate": 4.768199841395718e-05,
+      "loss": 1.3324,
+      "step": 8779
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1539249420166016,
+      "learning_rate": 4.7681734073486655e-05,
+      "loss": 1.3706,
+      "step": 8780
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.028362274169922,
+      "learning_rate": 4.768146973301613e-05,
+      "loss": 1.4629,
+      "step": 8781
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6638543605804443,
+      "learning_rate": 4.76812053925456e-05,
+      "loss": 1.3255,
+      "step": 8782
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.441721200942993,
+      "learning_rate": 4.7680941052075076e-05,
+      "loss": 1.5437,
+      "step": 8783
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0200841426849365,
+      "learning_rate": 4.768067671160455e-05,
+      "loss": 1.52,
+      "step": 8784
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5213662385940552,
+      "learning_rate": 4.768041237113402e-05,
+      "loss": 1.4471,
+      "step": 8785
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0746114253997803,
+      "learning_rate": 4.7680148030663496e-05,
+      "loss": 1.4417,
+      "step": 8786
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0043325424194336,
+      "learning_rate": 4.767988369019297e-05,
+      "loss": 1.4715,
+      "step": 8787
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6833045482635498,
+      "learning_rate": 4.7679619349722444e-05,
+      "loss": 1.4411,
+      "step": 8788
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.741995096206665,
+      "learning_rate": 4.767935500925192e-05,
+      "loss": 1.3455,
+      "step": 8789
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7207492589950562,
+      "learning_rate": 4.76790906687814e-05,
+      "loss": 1.2731,
+      "step": 8790
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6372286081314087,
+      "learning_rate": 4.7678826328310864e-05,
+      "loss": 1.4361,
+      "step": 8791
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4540146589279175,
+      "learning_rate": 4.7678561987840345e-05,
+      "loss": 1.2902,
+      "step": 8792
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6261627674102783,
+      "learning_rate": 4.767829764736981e-05,
+      "loss": 1.3808,
+      "step": 8793
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.734472632408142,
+      "learning_rate": 4.767803330689929e-05,
+      "loss": 1.6069,
+      "step": 8794
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.576294183731079,
+      "learning_rate": 4.7677768966428765e-05,
+      "loss": 1.424,
+      "step": 8795
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7728354930877686,
+      "learning_rate": 4.767750462595824e-05,
+      "loss": 1.4038,
+      "step": 8796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.833506464958191,
+      "learning_rate": 4.767724028548771e-05,
+      "loss": 1.5424,
+      "step": 8797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.681557297706604,
+      "learning_rate": 4.7676975945017186e-05,
+      "loss": 1.3361,
+      "step": 8798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6696927547454834,
+      "learning_rate": 4.767671160454666e-05,
+      "loss": 1.3571,
+      "step": 8799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0445597171783447,
+      "learning_rate": 4.767644726407613e-05,
+      "loss": 1.3821,
+      "step": 8800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6944050788879395,
+      "learning_rate": 4.767618292360561e-05,
+      "loss": 1.3807,
+      "step": 8801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.705519676208496,
+      "learning_rate": 4.767591858313508e-05,
+      "loss": 1.4139,
+      "step": 8802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.691151738166809,
+      "learning_rate": 4.7675654242664554e-05,
+      "loss": 1.3672,
+      "step": 8803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4173455238342285,
+      "learning_rate": 4.767538990219403e-05,
+      "loss": 1.6271,
+      "step": 8804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5584913492202759,
+      "learning_rate": 4.767512556172351e-05,
+      "loss": 1.3499,
+      "step": 8805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5816431045532227,
+      "learning_rate": 4.7674861221252975e-05,
+      "loss": 1.3141,
+      "step": 8806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9165116548538208,
+      "learning_rate": 4.7674596880782455e-05,
+      "loss": 1.5466,
+      "step": 8807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9461171627044678,
+      "learning_rate": 4.767433254031192e-05,
+      "loss": 1.397,
+      "step": 8808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.930853843688965,
+      "learning_rate": 4.7674068199841395e-05,
+      "loss": 1.3991,
+      "step": 8809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6137142181396484,
+      "learning_rate": 4.7673803859370876e-05,
+      "loss": 1.5367,
+      "step": 8810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6867111921310425,
+      "learning_rate": 4.767353951890034e-05,
+      "loss": 2.3969,
+      "step": 8811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.767849922180176,
+      "learning_rate": 4.767327517842982e-05,
+      "loss": 1.4118,
+      "step": 8812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6703988313674927,
+      "learning_rate": 4.767301083795929e-05,
+      "loss": 1.6681,
+      "step": 8813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.449528455734253,
+      "learning_rate": 4.767274649748877e-05,
+      "loss": 1.2843,
+      "step": 8814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8350625038146973,
+      "learning_rate": 4.7672482157018243e-05,
+      "loss": 1.3,
+      "step": 8815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9305238723754883,
+      "learning_rate": 4.767221781654772e-05,
+      "loss": 1.4411,
+      "step": 8816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2674453258514404,
+      "learning_rate": 4.767195347607719e-05,
+      "loss": 1.4943,
+      "step": 8817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6744717359542847,
+      "learning_rate": 4.7671689135606664e-05,
+      "loss": 1.2532,
+      "step": 8818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3411264419555664,
+      "learning_rate": 4.767142479513614e-05,
+      "loss": 1.2383,
+      "step": 8819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.148124933242798,
+      "learning_rate": 4.767116045466561e-05,
+      "loss": 1.4125,
+      "step": 8820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.619256854057312,
+      "learning_rate": 4.7670896114195085e-05,
+      "loss": 1.358,
+      "step": 8821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6589844226837158,
+      "learning_rate": 4.767063177372456e-05,
+      "loss": 1.5092,
+      "step": 8822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6548737287521362,
+      "learning_rate": 4.767036743325403e-05,
+      "loss": 1.376,
+      "step": 8823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2748780250549316,
+      "learning_rate": 4.7670103092783506e-05,
+      "loss": 1.5238,
+      "step": 8824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7316664457321167,
+      "learning_rate": 4.7669838752312986e-05,
+      "loss": 1.38,
+      "step": 8825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.786163330078125,
+      "learning_rate": 4.766957441184245e-05,
+      "loss": 1.3558,
+      "step": 8826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5207490921020508,
+      "learning_rate": 4.766931007137193e-05,
+      "loss": 1.2952,
+      "step": 8827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.860034465789795,
+      "learning_rate": 4.76690457309014e-05,
+      "loss": 1.4288,
+      "step": 8828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4277559518814087,
+      "learning_rate": 4.766878139043088e-05,
+      "loss": 2.211,
+      "step": 8829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7391276359558105,
+      "learning_rate": 4.7668517049960354e-05,
+      "loss": 1.561,
+      "step": 8830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6592984199523926,
+      "learning_rate": 4.766825270948983e-05,
+      "loss": 1.3778,
+      "step": 8831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5661979913711548,
+      "learning_rate": 4.76679883690193e-05,
+      "loss": 1.413,
+      "step": 8832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.02095890045166,
+      "learning_rate": 4.7667724028548775e-05,
+      "loss": 1.3529,
+      "step": 8833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7926307916641235,
+      "learning_rate": 4.766745968807825e-05,
+      "loss": 1.4157,
+      "step": 8834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5763660669326782,
+      "learning_rate": 4.766719534760772e-05,
+      "loss": 1.3036,
+      "step": 8835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6689561605453491,
+      "learning_rate": 4.7666931007137195e-05,
+      "loss": 1.249,
+      "step": 8836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.227633476257324,
+      "learning_rate": 4.766666666666667e-05,
+      "loss": 1.8272,
+      "step": 8837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7355221509933472,
+      "learning_rate": 4.766640232619614e-05,
+      "loss": 1.414,
+      "step": 8838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9351392984390259,
+      "learning_rate": 4.7666137985725616e-05,
+      "loss": 1.4159,
+      "step": 8839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4096810817718506,
+      "learning_rate": 4.7665873645255096e-05,
+      "loss": 1.3384,
+      "step": 8840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5776095390319824,
+      "learning_rate": 4.766560930478456e-05,
+      "loss": 1.3951,
+      "step": 8841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7804896831512451,
+      "learning_rate": 4.7665344964314043e-05,
+      "loss": 1.3917,
+      "step": 8842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3627190589904785,
+      "learning_rate": 4.766508062384351e-05,
+      "loss": 1.6223,
+      "step": 8843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5678342580795288,
+      "learning_rate": 4.766481628337299e-05,
+      "loss": 1.3747,
+      "step": 8844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8423203229904175,
+      "learning_rate": 4.7664551942902464e-05,
+      "loss": 1.3252,
+      "step": 8845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7820314168930054,
+      "learning_rate": 4.766428760243193e-05,
+      "loss": 1.4169,
+      "step": 8846
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7199146747589111,
+      "learning_rate": 4.766402326196141e-05,
+      "loss": 1.4638,
+      "step": 8847
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8143620491027832,
+      "learning_rate": 4.766375892149088e-05,
+      "loss": 1.3854,
+      "step": 8848
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1384456157684326,
+      "learning_rate": 4.766349458102036e-05,
+      "loss": 1.4619,
+      "step": 8849
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7704973220825195,
+      "learning_rate": 4.766323024054983e-05,
+      "loss": 1.4563,
+      "step": 8850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.749523401260376,
+      "learning_rate": 4.7662965900079306e-05,
+      "loss": 1.311,
+      "step": 8851
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.466542363166809,
+      "learning_rate": 4.766270155960878e-05,
+      "loss": 1.3183,
+      "step": 8852
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7187740802764893,
+      "learning_rate": 4.766243721913825e-05,
+      "loss": 1.3445,
+      "step": 8853
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.820098638534546,
+      "learning_rate": 4.7662172878667726e-05,
+      "loss": 1.4315,
+      "step": 8854
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.734941005706787,
+      "learning_rate": 4.76619085381972e-05,
+      "loss": 1.4376,
+      "step": 8855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.616091251373291,
+      "learning_rate": 4.7661644197726673e-05,
+      "loss": 1.3593,
+      "step": 8856
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6697971820831299,
+      "learning_rate": 4.766137985725615e-05,
+      "loss": 1.3821,
+      "step": 8857
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.418682098388672,
+      "learning_rate": 4.766111551678562e-05,
+      "loss": 1.5108,
+      "step": 8858
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4205408096313477,
+      "learning_rate": 4.7660851176315094e-05,
+      "loss": 1.4089,
+      "step": 8859
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.957651138305664,
+      "learning_rate": 4.7660586835844574e-05,
+      "loss": 1.4468,
+      "step": 8860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9936577081680298,
+      "learning_rate": 4.766032249537404e-05,
+      "loss": 1.469,
+      "step": 8861
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5157909393310547,
+      "learning_rate": 4.766005815490352e-05,
+      "loss": 1.3996,
+      "step": 8862
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.787728190422058,
+      "learning_rate": 4.765979381443299e-05,
+      "loss": 1.3358,
+      "step": 8863
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5595437288284302,
+      "learning_rate": 4.765952947396247e-05,
+      "loss": 1.2542,
+      "step": 8864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.472856283187866,
+      "learning_rate": 4.765926513349194e-05,
+      "loss": 1.6243,
+      "step": 8865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4466782808303833,
+      "learning_rate": 4.7659000793021416e-05,
+      "loss": 1.1424,
+      "step": 8866
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7375352382659912,
+      "learning_rate": 4.765873645255089e-05,
+      "loss": 1.4793,
+      "step": 8867
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7811883687973022,
+      "learning_rate": 4.765847211208036e-05,
+      "loss": 1.461,
+      "step": 8868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.248635172843933,
+      "learning_rate": 4.765820777160984e-05,
+      "loss": 1.2439,
+      "step": 8869
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8202751874923706,
+      "learning_rate": 4.765794343113931e-05,
+      "loss": 1.374,
+      "step": 8870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8673466444015503,
+      "learning_rate": 4.7657679090668784e-05,
+      "loss": 1.3292,
+      "step": 8871
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.076653242111206,
+      "learning_rate": 4.765741475019826e-05,
+      "loss": 1.4474,
+      "step": 8872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9462934732437134,
+      "learning_rate": 4.765715040972773e-05,
+      "loss": 1.6103,
+      "step": 8873
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3066778182983398,
+      "learning_rate": 4.7656886069257204e-05,
+      "loss": 1.2925,
+      "step": 8874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.15517520904541,
+      "learning_rate": 4.7656621728786685e-05,
+      "loss": 1.339,
+      "step": 8875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7306424379348755,
+      "learning_rate": 4.765635738831615e-05,
+      "loss": 1.4592,
+      "step": 8876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7769908905029297,
+      "learning_rate": 4.765609304784563e-05,
+      "loss": 1.2299,
+      "step": 8877
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8271660804748535,
+      "learning_rate": 4.76558287073751e-05,
+      "loss": 1.3105,
+      "step": 8878
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.479978084564209,
+      "learning_rate": 4.765556436690458e-05,
+      "loss": 1.1659,
+      "step": 8879
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.438313603401184,
+      "learning_rate": 4.7655300026434046e-05,
+      "loss": 1.3335,
+      "step": 8880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6201539039611816,
+      "learning_rate": 4.765503568596352e-05,
+      "loss": 1.2665,
+      "step": 8881
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0592947006225586,
+      "learning_rate": 4.7654771345493e-05,
+      "loss": 1.4153,
+      "step": 8882
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7160563468933105,
+      "learning_rate": 4.765450700502247e-05,
+      "loss": 1.4098,
+      "step": 8883
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7991353273391724,
+      "learning_rate": 4.765424266455195e-05,
+      "loss": 1.427,
+      "step": 8884
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7450101375579834,
+      "learning_rate": 4.7653978324081414e-05,
+      "loss": 1.421,
+      "step": 8885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7361621856689453,
+      "learning_rate": 4.7653713983610894e-05,
+      "loss": 1.3973,
+      "step": 8886
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0271124839782715,
+      "learning_rate": 4.765344964314037e-05,
+      "loss": 1.5508,
+      "step": 8887
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.734848976135254,
+      "learning_rate": 4.765318530266984e-05,
+      "loss": 1.3231,
+      "step": 8888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2452216148376465,
+      "learning_rate": 4.7652920962199315e-05,
+      "loss": 1.6212,
+      "step": 8889
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5995434522628784,
+      "learning_rate": 4.765265662172879e-05,
+      "loss": 1.6558,
+      "step": 8890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.136345863342285,
+      "learning_rate": 4.765239228125826e-05,
+      "loss": 2.5852,
+      "step": 8891
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9408358335494995,
+      "learning_rate": 4.7652127940787736e-05,
+      "loss": 1.4074,
+      "step": 8892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.097565174102783,
+      "learning_rate": 4.765186360031721e-05,
+      "loss": 1.2877,
+      "step": 8893
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8246904611587524,
+      "learning_rate": 4.765159925984668e-05,
+      "loss": 1.3593,
+      "step": 8894
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.698068618774414,
+      "learning_rate": 4.7651334919376156e-05,
+      "loss": 1.3422,
+      "step": 8895
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0375609397888184,
+      "learning_rate": 4.765107057890563e-05,
+      "loss": 1.3864,
+      "step": 8896
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7438709735870361,
+      "learning_rate": 4.765080623843511e-05,
+      "loss": 1.3431,
+      "step": 8897
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5779844522476196,
+      "learning_rate": 4.765054189796458e-05,
+      "loss": 1.3599,
+      "step": 8898
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7960368394851685,
+      "learning_rate": 4.765027755749406e-05,
+      "loss": 1.397,
+      "step": 8899
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4279130697250366,
+      "learning_rate": 4.7650013217023524e-05,
+      "loss": 1.3589,
+      "step": 8900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.274833917617798,
+      "learning_rate": 4.7649748876553004e-05,
+      "loss": 1.6021,
+      "step": 8901
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4068673849105835,
+      "learning_rate": 4.764948453608248e-05,
+      "loss": 1.0322,
+      "step": 8902
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1714487075805664,
+      "learning_rate": 4.764922019561195e-05,
+      "loss": 1.6532,
+      "step": 8903
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.426553726196289,
+      "learning_rate": 4.7648955855141425e-05,
+      "loss": 1.5063,
+      "step": 8904
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9092917442321777,
+      "learning_rate": 4.76486915146709e-05,
+      "loss": 1.4334,
+      "step": 8905
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4220924377441406,
+      "learning_rate": 4.764842717420037e-05,
+      "loss": 1.3555,
+      "step": 8906
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5334631204605103,
+      "learning_rate": 4.7648162833729846e-05,
+      "loss": 1.3912,
+      "step": 8907
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.105368137359619,
+      "learning_rate": 4.764789849325932e-05,
+      "loss": 1.3462,
+      "step": 8908
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.858144998550415,
+      "learning_rate": 4.764763415278879e-05,
+      "loss": 1.3521,
+      "step": 8909
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1620590686798096,
+      "learning_rate": 4.7647369812318267e-05,
+      "loss": 1.4057,
+      "step": 8910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.621993899345398,
+      "learning_rate": 4.764710547184774e-05,
+      "loss": 1.4303,
+      "step": 8911
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5416897535324097,
+      "learning_rate": 4.764684113137722e-05,
+      "loss": 1.5588,
+      "step": 8912
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.274407982826233,
+      "learning_rate": 4.764657679090669e-05,
+      "loss": 1.3648,
+      "step": 8913
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5078439712524414,
+      "learning_rate": 4.764631245043617e-05,
+      "loss": 1.3911,
+      "step": 8914
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.776833176612854,
+      "learning_rate": 4.7646048109965634e-05,
+      "loss": 1.3658,
+      "step": 8915
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5482736825942993,
+      "learning_rate": 4.7645783769495115e-05,
+      "loss": 1.3871,
+      "step": 8916
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.026121139526367,
+      "learning_rate": 4.764551942902459e-05,
+      "loss": 1.392,
+      "step": 8917
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7165565490722656,
+      "learning_rate": 4.7645255088554055e-05,
+      "loss": 1.3799,
+      "step": 8918
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.332402229309082,
+      "learning_rate": 4.7644990748083536e-05,
+      "loss": 1.4247,
+      "step": 8919
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.029031753540039,
+      "learning_rate": 4.7644726407613e-05,
+      "loss": 1.4954,
+      "step": 8920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.395517349243164,
+      "learning_rate": 4.764446206714248e-05,
+      "loss": 1.5413,
+      "step": 8921
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.453019142150879,
+      "learning_rate": 4.7644197726671956e-05,
+      "loss": 1.5068,
+      "step": 8922
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7994170188903809,
+      "learning_rate": 4.764393338620143e-05,
+      "loss": 1.313,
+      "step": 8923
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5007835626602173,
+      "learning_rate": 4.76436690457309e-05,
+      "loss": 1.341,
+      "step": 8924
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.738426685333252,
+      "learning_rate": 4.764340470526038e-05,
+      "loss": 1.2101,
+      "step": 8925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9244606494903564,
+      "learning_rate": 4.764314036478985e-05,
+      "loss": 1.4976,
+      "step": 8926
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7792484760284424,
+      "learning_rate": 4.7642876024319324e-05,
+      "loss": 1.4305,
+      "step": 8927
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5887722969055176,
+      "learning_rate": 4.76426116838488e-05,
+      "loss": 1.3335,
+      "step": 8928
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7735540866851807,
+      "learning_rate": 4.764234734337827e-05,
+      "loss": 1.4868,
+      "step": 8929
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8664311170578003,
+      "learning_rate": 4.7642083002907745e-05,
+      "loss": 1.3483,
+      "step": 8930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.717086672782898,
+      "learning_rate": 4.764181866243722e-05,
+      "loss": 1.3921,
+      "step": 8931
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.967501163482666,
+      "learning_rate": 4.76415543219667e-05,
+      "loss": 1.4188,
+      "step": 8932
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5075263977050781,
+      "learning_rate": 4.7641289981496166e-05,
+      "loss": 1.4412,
+      "step": 8933
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4971723556518555,
+      "learning_rate": 4.7641025641025646e-05,
+      "loss": 1.3414,
+      "step": 8934
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8181401491165161,
+      "learning_rate": 4.764076130055511e-05,
+      "loss": 1.2797,
+      "step": 8935
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5330218076705933,
+      "learning_rate": 4.764049696008459e-05,
+      "loss": 1.2423,
+      "step": 8936
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6738991737365723,
+      "learning_rate": 4.7640232619614067e-05,
+      "loss": 1.6886,
+      "step": 8937
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4870474338531494,
+      "learning_rate": 4.763996827914354e-05,
+      "loss": 1.27,
+      "step": 8938
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.291125535964966,
+      "learning_rate": 4.7639703938673014e-05,
+      "loss": 1.5632,
+      "step": 8939
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.068018913269043,
+      "learning_rate": 4.763943959820249e-05,
+      "loss": 1.461,
+      "step": 8940
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5139765739440918,
+      "learning_rate": 4.763917525773196e-05,
+      "loss": 1.2973,
+      "step": 8941
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.60450279712677,
+      "learning_rate": 4.7638910917261434e-05,
+      "loss": 1.3505,
+      "step": 8942
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7962478399276733,
+      "learning_rate": 4.763864657679091e-05,
+      "loss": 2.5358,
+      "step": 8943
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6768602132797241,
+      "learning_rate": 4.763838223632038e-05,
+      "loss": 1.4212,
+      "step": 8944
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3793882131576538,
+      "learning_rate": 4.7638117895849855e-05,
+      "loss": 1.362,
+      "step": 8945
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1660244464874268,
+      "learning_rate": 4.763785355537933e-05,
+      "loss": 1.5627,
+      "step": 8946
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8631048202514648,
+      "learning_rate": 4.763758921490881e-05,
+      "loss": 1.8611,
+      "step": 8947
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9864587783813477,
+      "learning_rate": 4.7637324874438276e-05,
+      "loss": 1.4935,
+      "step": 8948
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.212977886199951,
+      "learning_rate": 4.7637060533967756e-05,
+      "loss": 1.5663,
+      "step": 8949
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5395666360855103,
+      "learning_rate": 4.763679619349722e-05,
+      "loss": 1.327,
+      "step": 8950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8276047706604004,
+      "learning_rate": 4.76365318530267e-05,
+      "loss": 1.3477,
+      "step": 8951
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5159549713134766,
+      "learning_rate": 4.763626751255618e-05,
+      "loss": 1.2663,
+      "step": 8952
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8276435136795044,
+      "learning_rate": 4.7636003172085644e-05,
+      "loss": 1.63,
+      "step": 8953
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6677966117858887,
+      "learning_rate": 4.7635738831615124e-05,
+      "loss": 1.4305,
+      "step": 8954
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4312993288040161,
+      "learning_rate": 4.763547449114459e-05,
+      "loss": 1.3109,
+      "step": 8955
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.843767762184143,
+      "learning_rate": 4.763521015067407e-05,
+      "loss": 1.4062,
+      "step": 8956
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1852188110351562,
+      "learning_rate": 4.7634945810203545e-05,
+      "loss": 1.5365,
+      "step": 8957
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4997289180755615,
+      "learning_rate": 4.763468146973302e-05,
+      "loss": 1.3606,
+      "step": 8958
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.558872938156128,
+      "learning_rate": 4.763441712926249e-05,
+      "loss": 2.5839,
+      "step": 8959
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9028809070587158,
+      "learning_rate": 4.7634152788791965e-05,
+      "loss": 1.4158,
+      "step": 8960
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.761725902557373,
+      "learning_rate": 4.763388844832144e-05,
+      "loss": 1.381,
+      "step": 8961
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.790281057357788,
+      "learning_rate": 4.763362410785092e-05,
+      "loss": 1.4108,
+      "step": 8962
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0911128520965576,
+      "learning_rate": 4.7633359767380386e-05,
+      "loss": 1.3995,
+      "step": 8963
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3403854370117188,
+      "learning_rate": 4.763309542690986e-05,
+      "loss": 1.6466,
+      "step": 8964
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7212958335876465,
+      "learning_rate": 4.763283108643933e-05,
+      "loss": 1.5755,
+      "step": 8965
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7806376218795776,
+      "learning_rate": 4.763256674596881e-05,
+      "loss": 1.3874,
+      "step": 8966
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.833801507949829,
+      "learning_rate": 4.763230240549829e-05,
+      "loss": 1.5492,
+      "step": 8967
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.039108991622925,
+      "learning_rate": 4.7632038065027754e-05,
+      "loss": 1.4815,
+      "step": 8968
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.616143822669983,
+      "learning_rate": 4.7631773724557234e-05,
+      "loss": 1.3329,
+      "step": 8969
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.939974308013916,
+      "learning_rate": 4.76315093840867e-05,
+      "loss": 1.3761,
+      "step": 8970
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8188949823379517,
+      "learning_rate": 4.763124504361618e-05,
+      "loss": 1.4779,
+      "step": 8971
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9749559164047241,
+      "learning_rate": 4.7630980703145655e-05,
+      "loss": 1.4728,
+      "step": 8972
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6074692010879517,
+      "learning_rate": 4.763071636267513e-05,
+      "loss": 1.4624,
+      "step": 8973
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.722849726676941,
+      "learning_rate": 4.76304520222046e-05,
+      "loss": 1.5204,
+      "step": 8974
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.691962480545044,
+      "learning_rate": 4.7630187681734076e-05,
+      "loss": 1.3706,
+      "step": 8975
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2439491748809814,
+      "learning_rate": 4.762992334126355e-05,
+      "loss": 1.4272,
+      "step": 8976
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8006863594055176,
+      "learning_rate": 4.762965900079302e-05,
+      "loss": 1.4081,
+      "step": 8977
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8346649408340454,
+      "learning_rate": 4.7629394660322497e-05,
+      "loss": 1.2897,
+      "step": 8978
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9109160900115967,
+      "learning_rate": 4.762913031985197e-05,
+      "loss": 1.5429,
+      "step": 8979
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3859777450561523,
+      "learning_rate": 4.7628865979381444e-05,
+      "loss": 1.5267,
+      "step": 8980
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6669071912765503,
+      "learning_rate": 4.762860163891092e-05,
+      "loss": 1.4101,
+      "step": 8981
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5178179740905762,
+      "learning_rate": 4.76283372984404e-05,
+      "loss": 1.2581,
+      "step": 8982
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7902722358703613,
+      "learning_rate": 4.7628072957969864e-05,
+      "loss": 1.4351,
+      "step": 8983
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.638548493385315,
+      "learning_rate": 4.7627808617499345e-05,
+      "loss": 1.2405,
+      "step": 8984
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4860790967941284,
+      "learning_rate": 4.762754427702881e-05,
+      "loss": 1.2871,
+      "step": 8985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.102008819580078,
+      "learning_rate": 4.762727993655829e-05,
+      "loss": 1.4505,
+      "step": 8986
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6336359977722168,
+      "learning_rate": 4.7627015596087765e-05,
+      "loss": 1.345,
+      "step": 8987
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1341969966888428,
+      "learning_rate": 4.762675125561724e-05,
+      "loss": 1.465,
+      "step": 8988
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.197622537612915,
+      "learning_rate": 4.762648691514671e-05,
+      "loss": 1.4143,
+      "step": 8989
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5543136596679688,
+      "learning_rate": 4.762622257467618e-05,
+      "loss": 1.2649,
+      "step": 8990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6730550527572632,
+      "learning_rate": 4.762595823420566e-05,
+      "loss": 1.3751,
+      "step": 8991
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1925477981567383,
+      "learning_rate": 4.762569389373513e-05,
+      "loss": 1.5168,
+      "step": 8992
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.220149278640747,
+      "learning_rate": 4.762542955326461e-05,
+      "loss": 1.3968,
+      "step": 8993
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3541861772537231,
+      "learning_rate": 4.762516521279408e-05,
+      "loss": 1.2424,
+      "step": 8994
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8805592060089111,
+      "learning_rate": 4.7624900872323554e-05,
+      "loss": 1.4312,
+      "step": 8995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2351638078689575,
+      "learning_rate": 4.762463653185303e-05,
+      "loss": 2.4725,
+      "step": 8996
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1829874515533447,
+      "learning_rate": 4.762437219138251e-05,
+      "loss": 1.4833,
+      "step": 8997
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8072729110717773,
+      "learning_rate": 4.7624107850911975e-05,
+      "loss": 1.4558,
+      "step": 8998
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.126666784286499,
+      "learning_rate": 4.762384351044145e-05,
+      "loss": 1.4385,
+      "step": 8999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.335611581802368,
+      "learning_rate": 4.762357916997092e-05,
+      "loss": 1.4796,
+      "step": 9000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0694820880889893,
+      "learning_rate": 4.7623314829500395e-05,
+      "loss": 1.3949,
+      "step": 9001
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7130452394485474,
+      "learning_rate": 4.7623050489029876e-05,
+      "loss": 1.3999,
+      "step": 9002
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.480494260787964,
+      "learning_rate": 4.762278614855934e-05,
+      "loss": 1.5802,
+      "step": 9003
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.578932046890259,
+      "learning_rate": 4.762252180808882e-05,
+      "loss": 1.5875,
+      "step": 9004
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3905918598175049,
+      "learning_rate": 4.762225746761829e-05,
+      "loss": 1.7191,
+      "step": 9005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.755642056465149,
+      "learning_rate": 4.762199312714777e-05,
+      "loss": 1.4621,
+      "step": 9006
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.958344578742981,
+      "learning_rate": 4.7621728786677244e-05,
+      "loss": 1.336,
+      "step": 9007
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3412721157073975,
+      "learning_rate": 4.762146444620672e-05,
+      "loss": 1.3156,
+      "step": 9008
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9426705837249756,
+      "learning_rate": 4.762120010573619e-05,
+      "loss": 1.5077,
+      "step": 9009
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8149522542953491,
+      "learning_rate": 4.7620935765265664e-05,
+      "loss": 1.4219,
+      "step": 9010
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4914554357528687,
+      "learning_rate": 4.762067142479514e-05,
+      "loss": 1.3653,
+      "step": 9011
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6675137281417847,
+      "learning_rate": 4.762040708432461e-05,
+      "loss": 1.3344,
+      "step": 9012
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4882333278656006,
+      "learning_rate": 4.7620142743854085e-05,
+      "loss": 1.2585,
+      "step": 9013
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.773193597793579,
+      "learning_rate": 4.761987840338356e-05,
+      "loss": 1.3735,
+      "step": 9014
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4980000257492065,
+      "learning_rate": 4.761961406291303e-05,
+      "loss": 1.3242,
+      "step": 9015
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0485622882843018,
+      "learning_rate": 4.7619349722442506e-05,
+      "loss": 1.4849,
+      "step": 9016
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.671629548072815,
+      "learning_rate": 4.7619085381971986e-05,
+      "loss": 1.3158,
+      "step": 9017
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2543368339538574,
+      "learning_rate": 4.761882104150145e-05,
+      "loss": 1.526,
+      "step": 9018
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4445323944091797,
+      "learning_rate": 4.761855670103093e-05,
+      "loss": 1.3418,
+      "step": 9019
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6315207481384277,
+      "learning_rate": 4.76182923605604e-05,
+      "loss": 1.3841,
+      "step": 9020
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6070853471755981,
+      "learning_rate": 4.761802802008988e-05,
+      "loss": 1.5328,
+      "step": 9021
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0401618480682373,
+      "learning_rate": 4.7617763679619354e-05,
+      "loss": 1.5605,
+      "step": 9022
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.244370460510254,
+      "learning_rate": 4.761749933914883e-05,
+      "loss": 1.3934,
+      "step": 9023
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6320182085037231,
+      "learning_rate": 4.76172349986783e-05,
+      "loss": 1.6444,
+      "step": 9024
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7439243793487549,
+      "learning_rate": 4.761697065820777e-05,
+      "loss": 1.3876,
+      "step": 9025
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0930533409118652,
+      "learning_rate": 4.761670631773725e-05,
+      "loss": 1.3653,
+      "step": 9026
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7845321893692017,
+      "learning_rate": 4.761644197726672e-05,
+      "loss": 1.2718,
+      "step": 9027
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.708060622215271,
+      "learning_rate": 4.7616177636796195e-05,
+      "loss": 1.3162,
+      "step": 9028
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8239474296569824,
+      "learning_rate": 4.761591329632567e-05,
+      "loss": 1.4674,
+      "step": 9029
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6375623941421509,
+      "learning_rate": 4.761564895585514e-05,
+      "loss": 1.4079,
+      "step": 9030
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8922792673110962,
+      "learning_rate": 4.7615384615384616e-05,
+      "loss": 1.4605,
+      "step": 9031
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4331142902374268,
+      "learning_rate": 4.7615120274914096e-05,
+      "loss": 1.3815,
+      "step": 9032
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.021890163421631,
+      "learning_rate": 4.761485593444356e-05,
+      "loss": 1.3047,
+      "step": 9033
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7139333486557007,
+      "learning_rate": 4.7614591593973044e-05,
+      "loss": 2.4047,
+      "step": 9034
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7687175273895264,
+      "learning_rate": 4.761432725350251e-05,
+      "loss": 1.4321,
+      "step": 9035
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1140220165252686,
+      "learning_rate": 4.7614062913031984e-05,
+      "loss": 1.4478,
+      "step": 9036
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4684865474700928,
+      "learning_rate": 4.7613798572561464e-05,
+      "loss": 1.4468,
+      "step": 9037
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6712257862091064,
+      "learning_rate": 4.761353423209093e-05,
+      "loss": 2.4001,
+      "step": 9038
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8782401084899902,
+      "learning_rate": 4.761326989162041e-05,
+      "loss": 1.4133,
+      "step": 9039
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6382817029953003,
+      "learning_rate": 4.761300555114988e-05,
+      "loss": 1.2294,
+      "step": 9040
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0563952922821045,
+      "learning_rate": 4.761274121067936e-05,
+      "loss": 1.3862,
+      "step": 9041
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.83957576751709,
+      "learning_rate": 4.761247687020883e-05,
+      "loss": 1.5331,
+      "step": 9042
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5061168670654297,
+      "learning_rate": 4.7612212529738306e-05,
+      "loss": 1.3402,
+      "step": 9043
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6494309902191162,
+      "learning_rate": 4.761194818926778e-05,
+      "loss": 1.2829,
+      "step": 9044
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.521610975265503,
+      "learning_rate": 4.761168384879725e-05,
+      "loss": 1.4884,
+      "step": 9045
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7137765884399414,
+      "learning_rate": 4.7611419508326726e-05,
+      "loss": 1.4177,
+      "step": 9046
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.715190052986145,
+      "learning_rate": 4.76111551678562e-05,
+      "loss": 1.3915,
+      "step": 9047
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8109030723571777,
+      "learning_rate": 4.7610890827385674e-05,
+      "loss": 1.3699,
+      "step": 9048
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.010676145553589,
+      "learning_rate": 4.761062648691515e-05,
+      "loss": 1.4023,
+      "step": 9049
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.736821174621582,
+      "learning_rate": 4.761036214644462e-05,
+      "loss": 1.3371,
+      "step": 9050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7923836708068848,
+      "learning_rate": 4.7610097805974094e-05,
+      "loss": 1.5468,
+      "step": 9051
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.53668212890625,
+      "learning_rate": 4.7609833465503575e-05,
+      "loss": 1.2871,
+      "step": 9052
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.642210841178894,
+      "learning_rate": 4.760956912503304e-05,
+      "loss": 1.3289,
+      "step": 9053
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7015435695648193,
+      "learning_rate": 4.760930478456252e-05,
+      "loss": 1.45,
+      "step": 9054
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.969334602355957,
+      "learning_rate": 4.760904044409199e-05,
+      "loss": 1.3932,
+      "step": 9055
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9108244180679321,
+      "learning_rate": 4.760877610362147e-05,
+      "loss": 1.4674,
+      "step": 9056
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.036895275115967,
+      "learning_rate": 4.760851176315094e-05,
+      "loss": 1.5401,
+      "step": 9057
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.298698902130127,
+      "learning_rate": 4.7608247422680416e-05,
+      "loss": 1.5676,
+      "step": 9058
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0703043937683105,
+      "learning_rate": 4.760798308220989e-05,
+      "loss": 1.4224,
+      "step": 9059
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4382596015930176,
+      "learning_rate": 4.760771874173936e-05,
+      "loss": 1.4796,
+      "step": 9060
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7715386152267456,
+      "learning_rate": 4.760745440126884e-05,
+      "loss": 1.4802,
+      "step": 9061
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1508405208587646,
+      "learning_rate": 4.760719006079831e-05,
+      "loss": 1.5076,
+      "step": 9062
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2365589141845703,
+      "learning_rate": 4.7606925720327784e-05,
+      "loss": 1.5893,
+      "step": 9063
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.398604393005371,
+      "learning_rate": 4.760666137985726e-05,
+      "loss": 1.5568,
+      "step": 9064
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6684424877166748,
+      "learning_rate": 4.760639703938673e-05,
+      "loss": 1.3321,
+      "step": 9065
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7672300338745117,
+      "learning_rate": 4.7606132698916205e-05,
+      "loss": 1.5684,
+      "step": 9066
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6878684759140015,
+      "learning_rate": 4.7605868358445685e-05,
+      "loss": 1.3064,
+      "step": 9067
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6448842287063599,
+      "learning_rate": 4.760560401797515e-05,
+      "loss": 1.6656,
+      "step": 9068
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8488032817840576,
+      "learning_rate": 4.760533967750463e-05,
+      "loss": 1.363,
+      "step": 9069
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6627440452575684,
+      "learning_rate": 4.76050753370341e-05,
+      "loss": 1.3171,
+      "step": 9070
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.05847430229187,
+      "learning_rate": 4.760481099656357e-05,
+      "loss": 1.4993,
+      "step": 9071
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8791680335998535,
+      "learning_rate": 4.760454665609305e-05,
+      "loss": 1.3927,
+      "step": 9072
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0074453353881836,
+      "learning_rate": 4.760428231562252e-05,
+      "loss": 1.3946,
+      "step": 9073
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9270918369293213,
+      "learning_rate": 4.7604017975152e-05,
+      "loss": 1.3496,
+      "step": 9074
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6779592037200928,
+      "learning_rate": 4.760375363468147e-05,
+      "loss": 1.4133,
+      "step": 9075
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.633785367012024,
+      "learning_rate": 4.760348929421095e-05,
+      "loss": 1.2793,
+      "step": 9076
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7000585794448853,
+      "learning_rate": 4.760322495374042e-05,
+      "loss": 1.2677,
+      "step": 9077
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8139699697494507,
+      "learning_rate": 4.7602960613269894e-05,
+      "loss": 1.4497,
+      "step": 9078
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.15582275390625,
+      "learning_rate": 4.760269627279937e-05,
+      "loss": 1.1302,
+      "step": 9079
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.447908401489258,
+      "learning_rate": 4.760243193232884e-05,
+      "loss": 1.3651,
+      "step": 9080
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7497282028198242,
+      "learning_rate": 4.7602167591858315e-05,
+      "loss": 1.4642,
+      "step": 9081
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8857051134109497,
+      "learning_rate": 4.760190325138779e-05,
+      "loss": 1.4968,
+      "step": 9082
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.602142095565796,
+      "learning_rate": 4.760163891091726e-05,
+      "loss": 1.3139,
+      "step": 9083
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7625317573547363,
+      "learning_rate": 4.7601374570446736e-05,
+      "loss": 1.3977,
+      "step": 9084
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4607247114181519,
+      "learning_rate": 4.760111022997621e-05,
+      "loss": 1.4209,
+      "step": 9085
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.388629198074341,
+      "learning_rate": 4.760084588950568e-05,
+      "loss": 1.367,
+      "step": 9086
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9743410348892212,
+      "learning_rate": 4.760058154903516e-05,
+      "loss": 1.5032,
+      "step": 9087
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.694406032562256,
+      "learning_rate": 4.760031720856463e-05,
+      "loss": 1.4037,
+      "step": 9088
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2718276977539062,
+      "learning_rate": 4.760005286809411e-05,
+      "loss": 1.3469,
+      "step": 9089
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.026873826980591,
+      "learning_rate": 4.759978852762358e-05,
+      "loss": 2.2175,
+      "step": 9090
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8061810731887817,
+      "learning_rate": 4.759952418715306e-05,
+      "loss": 1.308,
+      "step": 9091
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8797938823699951,
+      "learning_rate": 4.759925984668253e-05,
+      "loss": 1.3948,
+      "step": 9092
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.330003499984741,
+      "learning_rate": 4.7598995506212005e-05,
+      "loss": 1.3416,
+      "step": 9093
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.967146396636963,
+      "learning_rate": 4.759873116574148e-05,
+      "loss": 1.392,
+      "step": 9094
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.382127046585083,
+      "learning_rate": 4.759846682527095e-05,
+      "loss": 1.3388,
+      "step": 9095
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9648306369781494,
+      "learning_rate": 4.7598202484800425e-05,
+      "loss": 1.6535,
+      "step": 9096
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5326277017593384,
+      "learning_rate": 4.75979381443299e-05,
+      "loss": 1.287,
+      "step": 9097
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1407101154327393,
+      "learning_rate": 4.759767380385937e-05,
+      "loss": 1.3947,
+      "step": 9098
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.926011085510254,
+      "learning_rate": 4.7597409463388846e-05,
+      "loss": 1.5964,
+      "step": 9099
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.251176118850708,
+      "learning_rate": 4.759714512291832e-05,
+      "loss": 1.3689,
+      "step": 9100
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.050088405609131,
+      "learning_rate": 4.759688078244779e-05,
+      "loss": 1.4852,
+      "step": 9101
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4712706804275513,
+      "learning_rate": 4.7596616441977274e-05,
+      "loss": 1.4086,
+      "step": 9102
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.841764211654663,
+      "learning_rate": 4.759635210150674e-05,
+      "loss": 1.4052,
+      "step": 9103
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9233795404434204,
+      "learning_rate": 4.759608776103622e-05,
+      "loss": 1.4145,
+      "step": 9104
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5960692167282104,
+      "learning_rate": 4.759582342056569e-05,
+      "loss": 1.4697,
+      "step": 9105
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.829012155532837,
+      "learning_rate": 4.759555908009517e-05,
+      "loss": 1.6442,
+      "step": 9106
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6072620153427124,
+      "learning_rate": 4.759529473962464e-05,
+      "loss": 2.1067,
+      "step": 9107
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5994335412979126,
+      "learning_rate": 4.759503039915411e-05,
+      "loss": 1.411,
+      "step": 9108
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.104534864425659,
+      "learning_rate": 4.759476605868359e-05,
+      "loss": 1.4572,
+      "step": 9109
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5995670557022095,
+      "learning_rate": 4.7594501718213055e-05,
+      "loss": 1.4498,
+      "step": 9110
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8087749481201172,
+      "learning_rate": 4.7594237377742536e-05,
+      "loss": 1.4092,
+      "step": 9111
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.502755045890808,
+      "learning_rate": 4.759397303727201e-05,
+      "loss": 1.42,
+      "step": 9112
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.972861409187317,
+      "learning_rate": 4.759370869680148e-05,
+      "loss": 1.4764,
+      "step": 9113
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.52084219455719,
+      "learning_rate": 4.7593444356330956e-05,
+      "loss": 1.3906,
+      "step": 9114
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7363629341125488,
+      "learning_rate": 4.759318001586043e-05,
+      "loss": 1.3876,
+      "step": 9115
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6464842557907104,
+      "learning_rate": 4.7592915675389904e-05,
+      "loss": 1.3706,
+      "step": 9116
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6084386110305786,
+      "learning_rate": 4.759265133491938e-05,
+      "loss": 1.3905,
+      "step": 9117
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.94711434841156,
+      "learning_rate": 4.759238699444885e-05,
+      "loss": 1.6766,
+      "step": 9118
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8544745445251465,
+      "learning_rate": 4.7592122653978324e-05,
+      "loss": 1.4829,
+      "step": 9119
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5783244371414185,
+      "learning_rate": 4.75918583135078e-05,
+      "loss": 1.252,
+      "step": 9120
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7524958848953247,
+      "learning_rate": 4.759159397303727e-05,
+      "loss": 1.5233,
+      "step": 9121
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8506735563278198,
+      "learning_rate": 4.759132963256675e-05,
+      "loss": 1.3668,
+      "step": 9122
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5367900133132935,
+      "learning_rate": 4.759106529209622e-05,
+      "loss": 1.4631,
+      "step": 9123
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7136187553405762,
+      "learning_rate": 4.75908009516257e-05,
+      "loss": 1.4131,
+      "step": 9124
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7853153944015503,
+      "learning_rate": 4.7590536611155166e-05,
+      "loss": 1.3556,
+      "step": 9125
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.754306435585022,
+      "learning_rate": 4.7590272270684646e-05,
+      "loss": 1.3904,
+      "step": 9126
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9295954704284668,
+      "learning_rate": 4.759000793021412e-05,
+      "loss": 1.3907,
+      "step": 9127
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.832277536392212,
+      "learning_rate": 4.758974358974359e-05,
+      "loss": 1.4058,
+      "step": 9128
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7784565687179565,
+      "learning_rate": 4.758947924927307e-05,
+      "loss": 1.4051,
+      "step": 9129
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2236281633377075,
+      "learning_rate": 4.758921490880254e-05,
+      "loss": 1.6061,
+      "step": 9130
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5484822988510132,
+      "learning_rate": 4.7588950568332014e-05,
+      "loss": 1.3236,
+      "step": 9131
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.673524022102356,
+      "learning_rate": 4.758868622786149e-05,
+      "loss": 1.6472,
+      "step": 9132
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.88094961643219,
+      "learning_rate": 4.758842188739096e-05,
+      "loss": 1.5471,
+      "step": 9133
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5611096620559692,
+      "learning_rate": 4.7588157546920435e-05,
+      "loss": 1.3704,
+      "step": 9134
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.680547833442688,
+      "learning_rate": 4.758789320644991e-05,
+      "loss": 1.3256,
+      "step": 9135
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4700430631637573,
+      "learning_rate": 4.758762886597938e-05,
+      "loss": 1.2805,
+      "step": 9136
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0165762901306152,
+      "learning_rate": 4.758736452550886e-05,
+      "loss": 1.5395,
+      "step": 9137
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7611507177352905,
+      "learning_rate": 4.758710018503833e-05,
+      "loss": 1.3963,
+      "step": 9138
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6739872694015503,
+      "learning_rate": 4.758683584456781e-05,
+      "loss": 1.2965,
+      "step": 9139
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5616159439086914,
+      "learning_rate": 4.7586571504097276e-05,
+      "loss": 1.4185,
+      "step": 9140
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.029881477355957,
+      "learning_rate": 4.7586307163626756e-05,
+      "loss": 1.4391,
+      "step": 9141
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9296560287475586,
+      "learning_rate": 4.758604282315623e-05,
+      "loss": 1.4991,
+      "step": 9142
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6420005559921265,
+      "learning_rate": 4.75857784826857e-05,
+      "loss": 1.2832,
+      "step": 9143
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8966120481491089,
+      "learning_rate": 4.758551414221518e-05,
+      "loss": 1.4375,
+      "step": 9144
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.786418080329895,
+      "learning_rate": 4.7585249801744644e-05,
+      "loss": 1.2895,
+      "step": 9145
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7625880241394043,
+      "learning_rate": 4.7584985461274124e-05,
+      "loss": 1.3928,
+      "step": 9146
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9475090503692627,
+      "learning_rate": 4.75847211208036e-05,
+      "loss": 1.2816,
+      "step": 9147
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.899342656135559,
+      "learning_rate": 4.758445678033307e-05,
+      "loss": 1.4267,
+      "step": 9148
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9349626302719116,
+      "learning_rate": 4.7584192439862545e-05,
+      "loss": 1.4194,
+      "step": 9149
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.027647018432617,
+      "learning_rate": 4.758392809939202e-05,
+      "loss": 1.4203,
+      "step": 9150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.030994415283203,
+      "learning_rate": 4.758366375892149e-05,
+      "loss": 1.4695,
+      "step": 9151
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4777127504348755,
+      "learning_rate": 4.758339941845097e-05,
+      "loss": 1.2451,
+      "step": 9152
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8967440128326416,
+      "learning_rate": 4.758313507798044e-05,
+      "loss": 1.7682,
+      "step": 9153
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4772274494171143,
+      "learning_rate": 4.758287073750991e-05,
+      "loss": 1.1899,
+      "step": 9154
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8271478414535522,
+      "learning_rate": 4.7582606397039386e-05,
+      "loss": 1.3983,
+      "step": 9155
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.885151386260986,
+      "learning_rate": 4.758234205656886e-05,
+      "loss": 1.7295,
+      "step": 9156
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.678471326828003,
+      "learning_rate": 4.758207771609834e-05,
+      "loss": 1.4584,
+      "step": 9157
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6660916805267334,
+      "learning_rate": 4.758181337562781e-05,
+      "loss": 3.8845,
+      "step": 9158
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.969872236251831,
+      "learning_rate": 4.758154903515729e-05,
+      "loss": 1.3933,
+      "step": 9159
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4359403848648071,
+      "learning_rate": 4.7581284694686754e-05,
+      "loss": 1.3314,
+      "step": 9160
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.036001443862915,
+      "learning_rate": 4.7581020354216235e-05,
+      "loss": 1.4313,
+      "step": 9161
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.880377173423767,
+      "learning_rate": 4.758075601374571e-05,
+      "loss": 1.3591,
+      "step": 9162
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7917331457138062,
+      "learning_rate": 4.758049167327518e-05,
+      "loss": 1.2851,
+      "step": 9163
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.125779390335083,
+      "learning_rate": 4.7580227332804655e-05,
+      "loss": 1.4484,
+      "step": 9164
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7482134103775024,
+      "learning_rate": 4.757996299233413e-05,
+      "loss": 1.4476,
+      "step": 9165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5866175889968872,
+      "learning_rate": 4.75796986518636e-05,
+      "loss": 1.3949,
+      "step": 9166
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6381494998931885,
+      "learning_rate": 4.7579434311393076e-05,
+      "loss": 1.3787,
+      "step": 9167
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.832493782043457,
+      "learning_rate": 4.757916997092255e-05,
+      "loss": 1.2601,
+      "step": 9168
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1128625869750977,
+      "learning_rate": 4.757890563045202e-05,
+      "loss": 1.5015,
+      "step": 9169
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6162272691726685,
+      "learning_rate": 4.75786412899815e-05,
+      "loss": 1.3882,
+      "step": 9170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5122071504592896,
+      "learning_rate": 4.757837694951097e-05,
+      "loss": 1.3768,
+      "step": 9171
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9214860200881958,
+      "learning_rate": 4.757811260904045e-05,
+      "loss": 1.553,
+      "step": 9172
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8788655996322632,
+      "learning_rate": 4.757784826856992e-05,
+      "loss": 1.3305,
+      "step": 9173
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5952041149139404,
+      "learning_rate": 4.75775839280994e-05,
+      "loss": 1.611,
+      "step": 9174
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.142489194869995,
+      "learning_rate": 4.7577319587628865e-05,
+      "loss": 1.5461,
+      "step": 9175
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8748588562011719,
+      "learning_rate": 4.7577055247158345e-05,
+      "loss": 1.3919,
+      "step": 9176
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.548450231552124,
+      "learning_rate": 4.757679090668782e-05,
+      "loss": 1.4031,
+      "step": 9177
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6491014957427979,
+      "learning_rate": 4.757652656621729e-05,
+      "loss": 1.4392,
+      "step": 9178
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4450337886810303,
+      "learning_rate": 4.7576262225746766e-05,
+      "loss": 1.5303,
+      "step": 9179
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6596710681915283,
+      "learning_rate": 4.757599788527623e-05,
+      "loss": 1.3775,
+      "step": 9180
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1316142082214355,
+      "learning_rate": 4.757573354480571e-05,
+      "loss": 1.4733,
+      "step": 9181
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.678294062614441,
+      "learning_rate": 4.7575469204335186e-05,
+      "loss": 1.3129,
+      "step": 9182
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8963394165039062,
+      "learning_rate": 4.757520486386466e-05,
+      "loss": 1.4037,
+      "step": 9183
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.431885838508606,
+      "learning_rate": 4.7574940523394133e-05,
+      "loss": 1.1353,
+      "step": 9184
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8566744327545166,
+      "learning_rate": 4.757467618292361e-05,
+      "loss": 1.3351,
+      "step": 9185
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8713080883026123,
+      "learning_rate": 4.757441184245308e-05,
+      "loss": 1.5138,
+      "step": 9186
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.18923020362854,
+      "learning_rate": 4.757414750198256e-05,
+      "loss": 1.4179,
+      "step": 9187
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8258330821990967,
+      "learning_rate": 4.757388316151203e-05,
+      "loss": 1.356,
+      "step": 9188
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0093564987182617,
+      "learning_rate": 4.75736188210415e-05,
+      "loss": 1.3626,
+      "step": 9189
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6786003112792969,
+      "learning_rate": 4.7573354480570975e-05,
+      "loss": 1.4592,
+      "step": 9190
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5968214273452759,
+      "learning_rate": 4.757309014010045e-05,
+      "loss": 1.2735,
+      "step": 9191
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8130325078964233,
+      "learning_rate": 4.757282579962993e-05,
+      "loss": 1.3901,
+      "step": 9192
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6145718097686768,
+      "learning_rate": 4.7572561459159396e-05,
+      "loss": 1.2959,
+      "step": 9193
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.89214026927948,
+      "learning_rate": 4.7572297118688876e-05,
+      "loss": 1.2815,
+      "step": 9194
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9980889558792114,
+      "learning_rate": 4.757203277821834e-05,
+      "loss": 1.3482,
+      "step": 9195
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.634369134902954,
+      "learning_rate": 4.757176843774782e-05,
+      "loss": 1.2589,
+      "step": 9196
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5587421655654907,
+      "learning_rate": 4.75715040972773e-05,
+      "loss": 1.3997,
+      "step": 9197
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.678231954574585,
+      "learning_rate": 4.757123975680677e-05,
+      "loss": 1.441,
+      "step": 9198
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1095354557037354,
+      "learning_rate": 4.7570975416336244e-05,
+      "loss": 2.2556,
+      "step": 9199
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9850962162017822,
+      "learning_rate": 4.757071107586572e-05,
+      "loss": 1.7266,
+      "step": 9200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1910321712493896,
+      "learning_rate": 4.757044673539519e-05,
+      "loss": 1.4565,
+      "step": 9201
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5926331281661987,
+      "learning_rate": 4.7570182394924665e-05,
+      "loss": 1.3027,
+      "step": 9202
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6590138673782349,
+      "learning_rate": 4.756991805445414e-05,
+      "loss": 1.3271,
+      "step": 9203
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6037979125976562,
+      "learning_rate": 4.756965371398361e-05,
+      "loss": 1.2811,
+      "step": 9204
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7153671979904175,
+      "learning_rate": 4.7569389373513085e-05,
+      "loss": 1.396,
+      "step": 9205
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.356539249420166,
+      "learning_rate": 4.756912503304256e-05,
+      "loss": 1.4981,
+      "step": 9206
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6886461973190308,
+      "learning_rate": 4.756886069257204e-05,
+      "loss": 1.3301,
+      "step": 9207
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8021210432052612,
+      "learning_rate": 4.7568596352101506e-05,
+      "loss": 1.3238,
+      "step": 9208
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.680417776107788,
+      "learning_rate": 4.7568332011630986e-05,
+      "loss": 1.3576,
+      "step": 9209
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4703364372253418,
+      "learning_rate": 4.756806767116045e-05,
+      "loss": 1.3547,
+      "step": 9210
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.872706413269043,
+      "learning_rate": 4.7567803330689933e-05,
+      "loss": 1.2134,
+      "step": 9211
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6670289039611816,
+      "learning_rate": 4.756753899021941e-05,
+      "loss": 1.4066,
+      "step": 9212
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0565154552459717,
+      "learning_rate": 4.756727464974888e-05,
+      "loss": 1.4893,
+      "step": 9213
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7285879850387573,
+      "learning_rate": 4.7567010309278354e-05,
+      "loss": 1.4781,
+      "step": 9214
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7658939361572266,
+      "learning_rate": 4.756674596880782e-05,
+      "loss": 1.4937,
+      "step": 9215
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8027653694152832,
+      "learning_rate": 4.75664816283373e-05,
+      "loss": 1.4972,
+      "step": 9216
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.850603699684143,
+      "learning_rate": 4.7566217287866775e-05,
+      "loss": 1.4619,
+      "step": 9217
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.86931574344635,
+      "learning_rate": 4.756595294739625e-05,
+      "loss": 1.4536,
+      "step": 9218
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5042001008987427,
+      "learning_rate": 4.756568860692572e-05,
+      "loss": 1.298,
+      "step": 9219
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.099210739135742,
+      "learning_rate": 4.7565424266455196e-05,
+      "loss": 1.4792,
+      "step": 9220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.589152455329895,
+      "learning_rate": 4.756515992598467e-05,
+      "loss": 1.2563,
+      "step": 9221
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5894793272018433,
+      "learning_rate": 4.756489558551415e-05,
+      "loss": 1.4091,
+      "step": 9222
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4234144687652588,
+      "learning_rate": 4.7564631245043616e-05,
+      "loss": 1.3595,
+      "step": 9223
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8324885368347168,
+      "learning_rate": 4.75643669045731e-05,
+      "loss": 1.4716,
+      "step": 9224
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1711862087249756,
+      "learning_rate": 4.7564102564102563e-05,
+      "loss": 1.3204,
+      "step": 9225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6392472982406616,
+      "learning_rate": 4.756383822363204e-05,
+      "loss": 1.3594,
+      "step": 9226
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.334646224975586,
+      "learning_rate": 4.756357388316152e-05,
+      "loss": 1.4336,
+      "step": 9227
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7742189168930054,
+      "learning_rate": 4.7563309542690984e-05,
+      "loss": 1.4827,
+      "step": 9228
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6558451652526855,
+      "learning_rate": 4.7563045202220464e-05,
+      "loss": 1.4039,
+      "step": 9229
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.312240719795227,
+      "learning_rate": 4.756278086174993e-05,
+      "loss": 1.2529,
+      "step": 9230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8161647319793701,
+      "learning_rate": 4.756251652127941e-05,
+      "loss": 1.423,
+      "step": 9231
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8281461000442505,
+      "learning_rate": 4.7562252180808885e-05,
+      "loss": 1.4881,
+      "step": 9232
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7154523134231567,
+      "learning_rate": 4.756198784033836e-05,
+      "loss": 1.2931,
+      "step": 9233
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9902095794677734,
+      "learning_rate": 4.756172349986783e-05,
+      "loss": 1.4982,
+      "step": 9234
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.666083335876465,
+      "learning_rate": 4.7561459159397306e-05,
+      "loss": 1.6093,
+      "step": 9235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2462010383605957,
+      "learning_rate": 4.756119481892678e-05,
+      "loss": 1.6124,
+      "step": 9236
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5937128067016602,
+      "learning_rate": 4.756093047845625e-05,
+      "loss": 1.325,
+      "step": 9237
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6166622638702393,
+      "learning_rate": 4.7560666137985727e-05,
+      "loss": 1.2985,
+      "step": 9238
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.207792043685913,
+      "learning_rate": 4.75604017975152e-05,
+      "loss": 1.4597,
+      "step": 9239
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.585115909576416,
+      "learning_rate": 4.7560137457044674e-05,
+      "loss": 1.3472,
+      "step": 9240
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.668380618095398,
+      "learning_rate": 4.755987311657415e-05,
+      "loss": 1.3057,
+      "step": 9241
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.078768730163574,
+      "learning_rate": 4.755960877610363e-05,
+      "loss": 1.3266,
+      "step": 9242
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5694531202316284,
+      "learning_rate": 4.7559344435633094e-05,
+      "loss": 1.4515,
+      "step": 9243
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.893362045288086,
+      "learning_rate": 4.7559080095162575e-05,
+      "loss": 1.8552,
+      "step": 9244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6100375652313232,
+      "learning_rate": 4.755881575469204e-05,
+      "loss": 1.4088,
+      "step": 9245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8235697746276855,
+      "learning_rate": 4.755855141422152e-05,
+      "loss": 1.5768,
+      "step": 9246
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.828650951385498,
+      "learning_rate": 4.7558287073750996e-05,
+      "loss": 1.5008,
+      "step": 9247
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9494625329971313,
+      "learning_rate": 4.755802273328047e-05,
+      "loss": 1.4687,
+      "step": 9248
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9190855026245117,
+      "learning_rate": 4.755775839280994e-05,
+      "loss": 1.5299,
+      "step": 9249
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5266938209533691,
+      "learning_rate": 4.7557494052339416e-05,
+      "loss": 1.3413,
+      "step": 9250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1407861709594727,
+      "learning_rate": 4.755722971186889e-05,
+      "loss": 2.0123,
+      "step": 9251
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0081899166107178,
+      "learning_rate": 4.755696537139836e-05,
+      "loss": 2.4917,
+      "step": 9252
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.92092764377594,
+      "learning_rate": 4.755670103092784e-05,
+      "loss": 1.4326,
+      "step": 9253
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8798227310180664,
+      "learning_rate": 4.755643669045731e-05,
+      "loss": 1.5115,
+      "step": 9254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.117338180541992,
+      "learning_rate": 4.7556172349986784e-05,
+      "loss": 1.4568,
+      "step": 9255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8938077688217163,
+      "learning_rate": 4.755590800951626e-05,
+      "loss": 1.4712,
+      "step": 9256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.628920316696167,
+      "learning_rate": 4.755564366904574e-05,
+      "loss": 1.4707,
+      "step": 9257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.933973550796509,
+      "learning_rate": 4.7555379328575205e-05,
+      "loss": 1.4846,
+      "step": 9258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.484084129333496,
+      "learning_rate": 4.7555114988104685e-05,
+      "loss": 1.4039,
+      "step": 9259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7708971500396729,
+      "learning_rate": 4.755485064763415e-05,
+      "loss": 1.3318,
+      "step": 9260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3645734786987305,
+      "learning_rate": 4.7554586307163626e-05,
+      "loss": 1.4062,
+      "step": 9261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.311372756958008,
+      "learning_rate": 4.7554321966693106e-05,
+      "loss": 1.5983,
+      "step": 9262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8725861310958862,
+      "learning_rate": 4.755405762622257e-05,
+      "loss": 1.5514,
+      "step": 9263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5547925233840942,
+      "learning_rate": 4.755379328575205e-05,
+      "loss": 1.4092,
+      "step": 9264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.01908278465271,
+      "learning_rate": 4.755352894528152e-05,
+      "loss": 1.4301,
+      "step": 9265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5564062595367432,
+      "learning_rate": 4.7553264604811e-05,
+      "loss": 1.3443,
+      "step": 9266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9880402088165283,
+      "learning_rate": 4.7553000264340474e-05,
+      "loss": 1.4477,
+      "step": 9267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9181843996047974,
+      "learning_rate": 4.755273592386995e-05,
+      "loss": 1.3947,
+      "step": 9268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8961763381958008,
+      "learning_rate": 4.755247158339942e-05,
+      "loss": 1.4364,
+      "step": 9269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5495457649230957,
+      "learning_rate": 4.7552207242928894e-05,
+      "loss": 1.4731,
+      "step": 9270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8900493383407593,
+      "learning_rate": 4.755194290245837e-05,
+      "loss": 1.2523,
+      "step": 9271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9647266864776611,
+      "learning_rate": 4.755167856198784e-05,
+      "loss": 1.5852,
+      "step": 9272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6082727909088135,
+      "learning_rate": 4.7551414221517315e-05,
+      "loss": 1.413,
+      "step": 9273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6017343997955322,
+      "learning_rate": 4.755114988104679e-05,
+      "loss": 1.5002,
+      "step": 9274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7348464727401733,
+      "learning_rate": 4.755088554057626e-05,
+      "loss": 1.4638,
+      "step": 9275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7402875423431396,
+      "learning_rate": 4.7550621200105736e-05,
+      "loss": 1.4698,
+      "step": 9276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9025189876556396,
+      "learning_rate": 4.7550356859635216e-05,
+      "loss": 1.343,
+      "step": 9277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.695616364479065,
+      "learning_rate": 4.755009251916468e-05,
+      "loss": 1.239,
+      "step": 9278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0013599395751953,
+      "learning_rate": 4.754982817869416e-05,
+      "loss": 1.3978,
+      "step": 9279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0084967613220215,
+      "learning_rate": 4.754956383822363e-05,
+      "loss": 1.4182,
+      "step": 9280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8684592247009277,
+      "learning_rate": 4.754929949775311e-05,
+      "loss": 1.3877,
+      "step": 9281
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8404357433319092,
+      "learning_rate": 4.7549035157282584e-05,
+      "loss": 1.3597,
+      "step": 9282
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9183529615402222,
+      "learning_rate": 4.754877081681206e-05,
+      "loss": 1.4391,
+      "step": 9283
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7493157386779785,
+      "learning_rate": 4.754850647634153e-05,
+      "loss": 1.431,
+      "step": 9284
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4806398153305054,
+      "learning_rate": 4.7548242135871005e-05,
+      "loss": 1.3986,
+      "step": 9285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6443982124328613,
+      "learning_rate": 4.754797779540048e-05,
+      "loss": 2.1322,
+      "step": 9286
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2386503219604492,
+      "learning_rate": 4.754771345492995e-05,
+      "loss": 1.2781,
+      "step": 9287
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3146493434906006,
+      "learning_rate": 4.7547449114459425e-05,
+      "loss": 1.4317,
+      "step": 9288
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5802383422851562,
+      "learning_rate": 4.75471847739889e-05,
+      "loss": 1.3845,
+      "step": 9289
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7818922996520996,
+      "learning_rate": 4.754692043351837e-05,
+      "loss": 1.5509,
+      "step": 9290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9549388885498047,
+      "learning_rate": 4.7546656093047846e-05,
+      "loss": 1.5975,
+      "step": 9291
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5244183540344238,
+      "learning_rate": 4.7546391752577327e-05,
+      "loss": 1.3115,
+      "step": 9292
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5872676372528076,
+      "learning_rate": 4.754612741210679e-05,
+      "loss": 1.3979,
+      "step": 9293
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8031818866729736,
+      "learning_rate": 4.7545863071636274e-05,
+      "loss": 1.5894,
+      "step": 9294
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9916690587997437,
+      "learning_rate": 4.754559873116574e-05,
+      "loss": 1.2603,
+      "step": 9295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.836033582687378,
+      "learning_rate": 4.754533439069522e-05,
+      "loss": 1.6467,
+      "step": 9296
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7422778606414795,
+      "learning_rate": 4.7545070050224694e-05,
+      "loss": 1.4782,
+      "step": 9297
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6339489221572876,
+      "learning_rate": 4.754480570975416e-05,
+      "loss": 1.4139,
+      "step": 9298
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0558626651763916,
+      "learning_rate": 4.754454136928364e-05,
+      "loss": 1.4807,
+      "step": 9299
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.248737335205078,
+      "learning_rate": 4.754427702881311e-05,
+      "loss": 1.4588,
+      "step": 9300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8408249616622925,
+      "learning_rate": 4.754401268834259e-05,
+      "loss": 1.3747,
+      "step": 9301
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.80805242061615,
+      "learning_rate": 4.754374834787206e-05,
+      "loss": 1.2794,
+      "step": 9302
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6038358211517334,
+      "learning_rate": 4.7543484007401536e-05,
+      "loss": 1.3738,
+      "step": 9303
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9587045907974243,
+      "learning_rate": 4.754321966693101e-05,
+      "loss": 1.5682,
+      "step": 9304
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.041146993637085,
+      "learning_rate": 4.754295532646048e-05,
+      "loss": 1.4713,
+      "step": 9305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6121693849563599,
+      "learning_rate": 4.7542690985989957e-05,
+      "loss": 1.3625,
+      "step": 9306
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8145604133605957,
+      "learning_rate": 4.754242664551943e-05,
+      "loss": 1.4682,
+      "step": 9307
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.835066795349121,
+      "learning_rate": 4.7542162305048904e-05,
+      "loss": 1.3473,
+      "step": 9308
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7143735885620117,
+      "learning_rate": 4.754189796457838e-05,
+      "loss": 1.4183,
+      "step": 9309
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9113972187042236,
+      "learning_rate": 4.754163362410785e-05,
+      "loss": 1.4804,
+      "step": 9310
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.494611144065857,
+      "learning_rate": 4.7541369283637324e-05,
+      "loss": 1.1928,
+      "step": 9311
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.115108013153076,
+      "learning_rate": 4.7541104943166805e-05,
+      "loss": 1.5334,
+      "step": 9312
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1007111072540283,
+      "learning_rate": 4.754084060269627e-05,
+      "loss": 1.3357,
+      "step": 9313
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.278604030609131,
+      "learning_rate": 4.754057626222575e-05,
+      "loss": 1.4917,
+      "step": 9314
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.96830677986145,
+      "learning_rate": 4.754031192175522e-05,
+      "loss": 1.4621,
+      "step": 9315
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8411921262741089,
+      "learning_rate": 4.75400475812847e-05,
+      "loss": 1.3397,
+      "step": 9316
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4436604976654053,
+      "learning_rate": 4.753978324081417e-05,
+      "loss": 1.1996,
+      "step": 9317
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3801651000976562,
+      "learning_rate": 4.7539518900343646e-05,
+      "loss": 1.4904,
+      "step": 9318
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0796058177948,
+      "learning_rate": 4.753925455987312e-05,
+      "loss": 1.4665,
+      "step": 9319
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1186864376068115,
+      "learning_rate": 4.753899021940259e-05,
+      "loss": 1.5769,
+      "step": 9320
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6088732481002808,
+      "learning_rate": 4.753872587893207e-05,
+      "loss": 1.5131,
+      "step": 9321
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7707006931304932,
+      "learning_rate": 4.753846153846154e-05,
+      "loss": 1.4356,
+      "step": 9322
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8502018451690674,
+      "learning_rate": 4.7538197197991014e-05,
+      "loss": 1.3779,
+      "step": 9323
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8909661769866943,
+      "learning_rate": 4.753793285752049e-05,
+      "loss": 1.3018,
+      "step": 9324
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5195730924606323,
+      "learning_rate": 4.753766851704996e-05,
+      "loss": 1.406,
+      "step": 9325
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8578627109527588,
+      "learning_rate": 4.7537404176579435e-05,
+      "loss": 1.4807,
+      "step": 9326
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3522658348083496,
+      "learning_rate": 4.7537139836108915e-05,
+      "loss": 1.5222,
+      "step": 9327
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.874488353729248,
+      "learning_rate": 4.753687549563838e-05,
+      "loss": 1.4589,
+      "step": 9328
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.991658091545105,
+      "learning_rate": 4.753661115516786e-05,
+      "loss": 1.5704,
+      "step": 9329
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.51870858669281,
+      "learning_rate": 4.753634681469733e-05,
+      "loss": 1.3776,
+      "step": 9330
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9963454008102417,
+      "learning_rate": 4.753608247422681e-05,
+      "loss": 1.5697,
+      "step": 9331
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.788605809211731,
+      "learning_rate": 4.753581813375628e-05,
+      "loss": 1.4259,
+      "step": 9332
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4492173194885254,
+      "learning_rate": 4.753555379328575e-05,
+      "loss": 1.3606,
+      "step": 9333
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.667729139328003,
+      "learning_rate": 4.753528945281523e-05,
+      "loss": 1.3463,
+      "step": 9334
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0945611000061035,
+      "learning_rate": 4.75350251123447e-05,
+      "loss": 1.4699,
+      "step": 9335
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5452959537506104,
+      "learning_rate": 4.753476077187418e-05,
+      "loss": 2.4435,
+      "step": 9336
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4081605672836304,
+      "learning_rate": 4.753449643140365e-05,
+      "loss": 1.286,
+      "step": 9337
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1978158950805664,
+      "learning_rate": 4.7534232090933124e-05,
+      "loss": 1.5218,
+      "step": 9338
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.086073875427246,
+      "learning_rate": 4.75339677504626e-05,
+      "loss": 1.379,
+      "step": 9339
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8825839757919312,
+      "learning_rate": 4.753370340999207e-05,
+      "loss": 1.3924,
+      "step": 9340
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6204500198364258,
+      "learning_rate": 4.7533439069521545e-05,
+      "loss": 1.3737,
+      "step": 9341
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5242244005203247,
+      "learning_rate": 4.7533174729051025e-05,
+      "loss": 1.3324,
+      "step": 9342
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3993796110153198,
+      "learning_rate": 4.753291038858049e-05,
+      "loss": 1.3866,
+      "step": 9343
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8230681419372559,
+      "learning_rate": 4.7532646048109966e-05,
+      "loss": 1.4502,
+      "step": 9344
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 23.40035057067871,
+      "learning_rate": 4.753238170763944e-05,
+      "loss": 1.6983,
+      "step": 9345
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6026116609573364,
+      "learning_rate": 4.753211736716891e-05,
+      "loss": 1.4804,
+      "step": 9346
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5547354221343994,
+      "learning_rate": 4.753185302669839e-05,
+      "loss": 1.4048,
+      "step": 9347
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7814797163009644,
+      "learning_rate": 4.753158868622786e-05,
+      "loss": 1.4822,
+      "step": 9348
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.786340594291687,
+      "learning_rate": 4.753132434575734e-05,
+      "loss": 1.3266,
+      "step": 9349
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8533440828323364,
+      "learning_rate": 4.753106000528681e-05,
+      "loss": 1.2908,
+      "step": 9350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7026493549346924,
+      "learning_rate": 4.753079566481629e-05,
+      "loss": 1.3752,
+      "step": 9351
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2413578033447266,
+      "learning_rate": 4.753053132434576e-05,
+      "loss": 1.4723,
+      "step": 9352
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9844321012496948,
+      "learning_rate": 4.7530266983875235e-05,
+      "loss": 1.376,
+      "step": 9353
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8855922222137451,
+      "learning_rate": 4.753000264340471e-05,
+      "loss": 1.3992,
+      "step": 9354
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6247828006744385,
+      "learning_rate": 4.752973830293418e-05,
+      "loss": 2.0065,
+      "step": 9355
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6472744941711426,
+      "learning_rate": 4.7529473962463655e-05,
+      "loss": 1.3093,
+      "step": 9356
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.783633828163147,
+      "learning_rate": 4.752920962199313e-05,
+      "loss": 1.3339,
+      "step": 9357
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8719042539596558,
+      "learning_rate": 4.75289452815226e-05,
+      "loss": 1.5017,
+      "step": 9358
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.074244499206543,
+      "learning_rate": 4.7528680941052076e-05,
+      "loss": 1.5386,
+      "step": 9359
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.306819200515747,
+      "learning_rate": 4.752841660058155e-05,
+      "loss": 1.3513,
+      "step": 9360
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5728217363357544,
+      "learning_rate": 4.752815226011102e-05,
+      "loss": 1.236,
+      "step": 9361
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7597039937973022,
+      "learning_rate": 4.7527887919640504e-05,
+      "loss": 1.2566,
+      "step": 9362
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5851185321807861,
+      "learning_rate": 4.752762357916997e-05,
+      "loss": 1.3319,
+      "step": 9363
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4234153032302856,
+      "learning_rate": 4.752735923869945e-05,
+      "loss": 1.728,
+      "step": 9364
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.624635100364685,
+      "learning_rate": 4.752709489822892e-05,
+      "loss": 1.3671,
+      "step": 9365
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7403417825698853,
+      "learning_rate": 4.75268305577584e-05,
+      "loss": 1.3958,
+      "step": 9366
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9907394647598267,
+      "learning_rate": 4.752656621728787e-05,
+      "loss": 1.419,
+      "step": 9367
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.743443012237549,
+      "learning_rate": 4.7526301876817345e-05,
+      "loss": 1.5492,
+      "step": 9368
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.766688585281372,
+      "learning_rate": 4.752603753634682e-05,
+      "loss": 1.4566,
+      "step": 9369
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7847623825073242,
+      "learning_rate": 4.7525773195876285e-05,
+      "loss": 1.344,
+      "step": 9370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0143508911132812,
+      "learning_rate": 4.7525508855405766e-05,
+      "loss": 1.5017,
+      "step": 9371
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1949875354766846,
+      "learning_rate": 4.752524451493524e-05,
+      "loss": 1.4218,
+      "step": 9372
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5716652870178223,
+      "learning_rate": 4.752498017446471e-05,
+      "loss": 1.3955,
+      "step": 9373
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.638537883758545,
+      "learning_rate": 4.7524715833994186e-05,
+      "loss": 1.46,
+      "step": 9374
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0557453632354736,
+      "learning_rate": 4.752445149352366e-05,
+      "loss": 1.4887,
+      "step": 9375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.001035213470459,
+      "learning_rate": 4.7524187153053134e-05,
+      "loss": 1.4043,
+      "step": 9376
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.619336724281311,
+      "learning_rate": 4.7523922812582614e-05,
+      "loss": 1.2475,
+      "step": 9377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4386377334594727,
+      "learning_rate": 4.752365847211208e-05,
+      "loss": 1.5282,
+      "step": 9378
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8144317865371704,
+      "learning_rate": 4.7523394131641554e-05,
+      "loss": 1.3442,
+      "step": 9379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4788798093795776,
+      "learning_rate": 4.752312979117103e-05,
+      "loss": 1.5056,
+      "step": 9380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.862346887588501,
+      "learning_rate": 4.75228654507005e-05,
+      "loss": 1.5671,
+      "step": 9381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4330154657363892,
+      "learning_rate": 4.752260111022998e-05,
+      "loss": 1.3334,
+      "step": 9382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.214519500732422,
+      "learning_rate": 4.752233676975945e-05,
+      "loss": 1.2763,
+      "step": 9383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4029558897018433,
+      "learning_rate": 4.752207242928893e-05,
+      "loss": 1.3543,
+      "step": 9384
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0710651874542236,
+      "learning_rate": 4.7521808088818396e-05,
+      "loss": 1.3894,
+      "step": 9385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8245720863342285,
+      "learning_rate": 4.7521543748347876e-05,
+      "loss": 1.3771,
+      "step": 9386
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.011970281600952,
+      "learning_rate": 4.752127940787735e-05,
+      "loss": 1.4784,
+      "step": 9387
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5219391584396362,
+      "learning_rate": 4.752101506740682e-05,
+      "loss": 1.2669,
+      "step": 9388
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.795090675354004,
+      "learning_rate": 4.75207507269363e-05,
+      "loss": 1.4298,
+      "step": 9389
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7639623880386353,
+      "learning_rate": 4.752048638646577e-05,
+      "loss": 1.382,
+      "step": 9390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.039462089538574,
+      "learning_rate": 4.7520222045995244e-05,
+      "loss": 1.817,
+      "step": 9391
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4544917345046997,
+      "learning_rate": 4.751995770552472e-05,
+      "loss": 1.3241,
+      "step": 9392
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6890759468078613,
+      "learning_rate": 4.751969336505419e-05,
+      "loss": 1.3672,
+      "step": 9393
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5964930057525635,
+      "learning_rate": 4.7519429024583665e-05,
+      "loss": 1.3152,
+      "step": 9394
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.661321997642517,
+      "learning_rate": 4.751916468411314e-05,
+      "loss": 1.3531,
+      "step": 9395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5720030069351196,
+      "learning_rate": 4.751890034364261e-05,
+      "loss": 1.2643,
+      "step": 9396
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8673975467681885,
+      "learning_rate": 4.751863600317209e-05,
+      "loss": 1.5165,
+      "step": 9397
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7456116676330566,
+      "learning_rate": 4.751837166270156e-05,
+      "loss": 1.3168,
+      "step": 9398
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8333581686019897,
+      "learning_rate": 4.751810732223104e-05,
+      "loss": 1.3569,
+      "step": 9399
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.074610471725464,
+      "learning_rate": 4.7517842981760506e-05,
+      "loss": 1.4699,
+      "step": 9400
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9906558990478516,
+      "learning_rate": 4.7517578641289986e-05,
+      "loss": 1.438,
+      "step": 9401
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5325995683670044,
+      "learning_rate": 4.751731430081946e-05,
+      "loss": 1.2982,
+      "step": 9402
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4033921957015991,
+      "learning_rate": 4.7517049960348934e-05,
+      "loss": 1.2594,
+      "step": 9403
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.767665147781372,
+      "learning_rate": 4.751678561987841e-05,
+      "loss": 1.4492,
+      "step": 9404
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.524986743927002,
+      "learning_rate": 4.751652127940788e-05,
+      "loss": 2.3191,
+      "step": 9405
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.546691417694092,
+      "learning_rate": 4.7516256938937354e-05,
+      "loss": 1.4667,
+      "step": 9406
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.631300926208496,
+      "learning_rate": 4.751599259846683e-05,
+      "loss": 1.4469,
+      "step": 9407
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3069524765014648,
+      "learning_rate": 4.75157282579963e-05,
+      "loss": 1.3236,
+      "step": 9408
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.404860019683838,
+      "learning_rate": 4.7515463917525775e-05,
+      "loss": 1.3908,
+      "step": 9409
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.564683437347412,
+      "learning_rate": 4.751519957705525e-05,
+      "loss": 1.4267,
+      "step": 9410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6661134958267212,
+      "learning_rate": 4.751493523658472e-05,
+      "loss": 1.6058,
+      "step": 9411
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.706478476524353,
+      "learning_rate": 4.75146708961142e-05,
+      "loss": 1.3954,
+      "step": 9412
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.18823504447937,
+      "learning_rate": 4.751440655564367e-05,
+      "loss": 1.5405,
+      "step": 9413
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2216243743896484,
+      "learning_rate": 4.751414221517315e-05,
+      "loss": 1.4955,
+      "step": 9414
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8020392656326294,
+      "learning_rate": 4.7513877874702616e-05,
+      "loss": 1.3673,
+      "step": 9415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.168087959289551,
+      "learning_rate": 4.751361353423209e-05,
+      "loss": 1.7881,
+      "step": 9416
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9837045669555664,
+      "learning_rate": 4.751334919376157e-05,
+      "loss": 1.3713,
+      "step": 9417
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8064786195755005,
+      "learning_rate": 4.751308485329104e-05,
+      "loss": 1.4657,
+      "step": 9418
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.741943359375,
+      "learning_rate": 4.751282051282052e-05,
+      "loss": 1.5576,
+      "step": 9419
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3895407915115356,
+      "learning_rate": 4.7512556172349984e-05,
+      "loss": 1.4002,
+      "step": 9420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.922153115272522,
+      "learning_rate": 4.7512291831879465e-05,
+      "loss": 1.4742,
+      "step": 9421
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7440862655639648,
+      "learning_rate": 4.751202749140894e-05,
+      "loss": 1.3694,
+      "step": 9422
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6284270286560059,
+      "learning_rate": 4.751176315093841e-05,
+      "loss": 1.3367,
+      "step": 9423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1594154834747314,
+      "learning_rate": 4.7511498810467885e-05,
+      "loss": 1.508,
+      "step": 9424
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5122591257095337,
+      "learning_rate": 4.751123446999736e-05,
+      "loss": 1.7695,
+      "step": 9425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1443333625793457,
+      "learning_rate": 4.751097012952683e-05,
+      "loss": 1.4192,
+      "step": 9426
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7151331901550293,
+      "learning_rate": 4.7510705789056306e-05,
+      "loss": 1.5281,
+      "step": 9427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.401176691055298,
+      "learning_rate": 4.751044144858578e-05,
+      "loss": 1.4412,
+      "step": 9428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9985239505767822,
+      "learning_rate": 4.751017710811525e-05,
+      "loss": 1.5161,
+      "step": 9429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5840246677398682,
+      "learning_rate": 4.750991276764473e-05,
+      "loss": 1.2627,
+      "step": 9430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2833901643753052,
+      "learning_rate": 4.75096484271742e-05,
+      "loss": 1.2622,
+      "step": 9431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5398446321487427,
+      "learning_rate": 4.750938408670368e-05,
+      "loss": 1.3764,
+      "step": 9432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9153366088867188,
+      "learning_rate": 4.750911974623315e-05,
+      "loss": 1.4872,
+      "step": 9433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2648298740386963,
+      "learning_rate": 4.750885540576263e-05,
+      "loss": 1.6379,
+      "step": 9434
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9750510454177856,
+      "learning_rate": 4.7508591065292095e-05,
+      "loss": 1.4476,
+      "step": 9435
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8929686546325684,
+      "learning_rate": 4.7508326724821575e-05,
+      "loss": 1.3921,
+      "step": 9436
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6870394945144653,
+      "learning_rate": 4.750806238435105e-05,
+      "loss": 1.3202,
+      "step": 9437
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.425873041152954,
+      "learning_rate": 4.750779804388052e-05,
+      "loss": 1.3838,
+      "step": 9438
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0570571422576904,
+      "learning_rate": 4.7507533703409996e-05,
+      "loss": 1.3201,
+      "step": 9439
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.911962866783142,
+      "learning_rate": 4.750726936293947e-05,
+      "loss": 1.4438,
+      "step": 9440
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.811096668243408,
+      "learning_rate": 4.750700502246894e-05,
+      "loss": 1.6797,
+      "step": 9441
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6868081092834473,
+      "learning_rate": 4.7506740681998416e-05,
+      "loss": 1.4131,
+      "step": 9442
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8638981580734253,
+      "learning_rate": 4.750647634152789e-05,
+      "loss": 1.4713,
+      "step": 9443
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9259730577468872,
+      "learning_rate": 4.7506212001057364e-05,
+      "loss": 1.2969,
+      "step": 9444
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0648183822631836,
+      "learning_rate": 4.750594766058684e-05,
+      "loss": 1.6301,
+      "step": 9445
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3877663612365723,
+      "learning_rate": 4.750568332011631e-05,
+      "loss": 1.2731,
+      "step": 9446
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9492071866989136,
+      "learning_rate": 4.750541897964579e-05,
+      "loss": 1.5526,
+      "step": 9447
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4016951322555542,
+      "learning_rate": 4.750515463917526e-05,
+      "loss": 1.8785,
+      "step": 9448
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.163994789123535,
+      "learning_rate": 4.750489029870474e-05,
+      "loss": 1.8181,
+      "step": 9449
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.535757064819336,
+      "learning_rate": 4.7504625958234205e-05,
+      "loss": 1.4878,
+      "step": 9450
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.932408094406128,
+      "learning_rate": 4.750436161776368e-05,
+      "loss": 1.8521,
+      "step": 9451
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4106738567352295,
+      "learning_rate": 4.750409727729316e-05,
+      "loss": 1.2964,
+      "step": 9452
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5012328624725342,
+      "learning_rate": 4.7503832936822626e-05,
+      "loss": 1.3089,
+      "step": 9453
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.199774980545044,
+      "learning_rate": 4.7503568596352106e-05,
+      "loss": 1.4125,
+      "step": 9454
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6588385105133057,
+      "learning_rate": 4.750330425588157e-05,
+      "loss": 2.5396,
+      "step": 9455
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0497748851776123,
+      "learning_rate": 4.750303991541105e-05,
+      "loss": 1.4803,
+      "step": 9456
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6446107625961304,
+      "learning_rate": 4.750277557494053e-05,
+      "loss": 1.3735,
+      "step": 9457
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.438251256942749,
+      "learning_rate": 4.750251123447e-05,
+      "loss": 1.2762,
+      "step": 9458
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8752504587173462,
+      "learning_rate": 4.7502246893999474e-05,
+      "loss": 1.3078,
+      "step": 9459
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8747206926345825,
+      "learning_rate": 4.750198255352895e-05,
+      "loss": 1.5193,
+      "step": 9460
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.285569667816162,
+      "learning_rate": 4.750171821305842e-05,
+      "loss": 1.5027,
+      "step": 9461
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6302381753921509,
+      "learning_rate": 4.7501453872587895e-05,
+      "loss": 1.447,
+      "step": 9462
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9532657861709595,
+      "learning_rate": 4.750118953211737e-05,
+      "loss": 1.5768,
+      "step": 9463
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4246900081634521,
+      "learning_rate": 4.750092519164684e-05,
+      "loss": 1.2117,
+      "step": 9464
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1835548877716064,
+      "learning_rate": 4.7500660851176315e-05,
+      "loss": 1.3838,
+      "step": 9465
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3932793140411377,
+      "learning_rate": 4.750039651070579e-05,
+      "loss": 1.3059,
+      "step": 9466
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.580044150352478,
+      "learning_rate": 4.750013217023527e-05,
+      "loss": 1.4682,
+      "step": 9467
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5151705741882324,
+      "learning_rate": 4.7499867829764736e-05,
+      "loss": 1.3379,
+      "step": 9468
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.961941123008728,
+      "learning_rate": 4.7499603489294216e-05,
+      "loss": 1.4176,
+      "step": 9469
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.685718297958374,
+      "learning_rate": 4.749933914882368e-05,
+      "loss": 1.5119,
+      "step": 9470
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7308592796325684,
+      "learning_rate": 4.7499074808353164e-05,
+      "loss": 1.514,
+      "step": 9471
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5631706714630127,
+      "learning_rate": 4.749881046788264e-05,
+      "loss": 1.9733,
+      "step": 9472
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.844307541847229,
+      "learning_rate": 4.749854612741211e-05,
+      "loss": 1.4649,
+      "step": 9473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5675874948501587,
+      "learning_rate": 4.7498281786941584e-05,
+      "loss": 1.293,
+      "step": 9474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4467843770980835,
+      "learning_rate": 4.749801744647106e-05,
+      "loss": 1.4072,
+      "step": 9475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1486830711364746,
+      "learning_rate": 4.749775310600053e-05,
+      "loss": 1.4031,
+      "step": 9476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.613222122192383,
+      "learning_rate": 4.7497488765530005e-05,
+      "loss": 1.4075,
+      "step": 9477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.652279257774353,
+      "learning_rate": 4.749722442505948e-05,
+      "loss": 1.3753,
+      "step": 9478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.761998176574707,
+      "learning_rate": 4.749696008458895e-05,
+      "loss": 1.3951,
+      "step": 9479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7789430618286133,
+      "learning_rate": 4.7496695744118426e-05,
+      "loss": 1.654,
+      "step": 9480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9385244846343994,
+      "learning_rate": 4.74964314036479e-05,
+      "loss": 1.6448,
+      "step": 9481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4804562330245972,
+      "learning_rate": 4.749616706317738e-05,
+      "loss": 1.2008,
+      "step": 9482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.090029239654541,
+      "learning_rate": 4.7495902722706846e-05,
+      "loss": 1.595,
+      "step": 9483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9683094024658203,
+      "learning_rate": 4.749563838223633e-05,
+      "loss": 1.3659,
+      "step": 9484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8956369161605835,
+      "learning_rate": 4.7495374041765793e-05,
+      "loss": 1.4131,
+      "step": 9485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2543082237243652,
+      "learning_rate": 4.7495109701295274e-05,
+      "loss": 1.4591,
+      "step": 9486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7661689519882202,
+      "learning_rate": 4.749484536082475e-05,
+      "loss": 1.421,
+      "step": 9487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7193081378936768,
+      "learning_rate": 4.7494581020354214e-05,
+      "loss": 1.7715,
+      "step": 9488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5088932514190674,
+      "learning_rate": 4.7494316679883695e-05,
+      "loss": 1.3298,
+      "step": 9489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5224289894104004,
+      "learning_rate": 4.749405233941316e-05,
+      "loss": 1.4199,
+      "step": 9490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7588645219802856,
+      "learning_rate": 4.749378799894264e-05,
+      "loss": 1.4605,
+      "step": 9491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6421970129013062,
+      "learning_rate": 4.7493523658472115e-05,
+      "loss": 1.2706,
+      "step": 9492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5389411449432373,
+      "learning_rate": 4.749325931800159e-05,
+      "loss": 1.4959,
+      "step": 9493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6330242156982422,
+      "learning_rate": 4.749299497753106e-05,
+      "loss": 1.2832,
+      "step": 9494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9392796754837036,
+      "learning_rate": 4.7492730637060536e-05,
+      "loss": 1.3253,
+      "step": 9495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7058141231536865,
+      "learning_rate": 4.749246629659001e-05,
+      "loss": 1.3976,
+      "step": 9496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.452347755432129,
+      "learning_rate": 4.749220195611948e-05,
+      "loss": 1.4241,
+      "step": 9497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.805414080619812,
+      "learning_rate": 4.749193761564896e-05,
+      "loss": 1.3176,
+      "step": 9498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.865996241569519,
+      "learning_rate": 4.749167327517843e-05,
+      "loss": 1.4899,
+      "step": 9499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9835844039916992,
+      "learning_rate": 4.7491408934707904e-05,
+      "loss": 1.4665,
+      "step": 9500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5872269868850708,
+      "learning_rate": 4.749114459423738e-05,
+      "loss": 1.5955,
+      "step": 9501
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4706368446350098,
+      "learning_rate": 4.749088025376686e-05,
+      "loss": 1.3459,
+      "step": 9502
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.489326000213623,
+      "learning_rate": 4.7490615913296325e-05,
+      "loss": 2.0728,
+      "step": 9503
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7418092489242554,
+      "learning_rate": 4.7490351572825805e-05,
+      "loss": 1.2388,
+      "step": 9504
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4211612939834595,
+      "learning_rate": 4.749008723235527e-05,
+      "loss": 1.2856,
+      "step": 9505
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.586256980895996,
+      "learning_rate": 4.748982289188475e-05,
+      "loss": 1.4513,
+      "step": 9506
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.459160566329956,
+      "learning_rate": 4.7489558551414226e-05,
+      "loss": 1.3816,
+      "step": 9507
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6262542009353638,
+      "learning_rate": 4.74892942109437e-05,
+      "loss": 1.3675,
+      "step": 9508
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8451933860778809,
+      "learning_rate": 4.748902987047317e-05,
+      "loss": 1.3622,
+      "step": 9509
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.51716947555542,
+      "learning_rate": 4.7488765530002646e-05,
+      "loss": 1.3329,
+      "step": 9510
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8425272703170776,
+      "learning_rate": 4.748850118953212e-05,
+      "loss": 1.4545,
+      "step": 9511
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6701304912567139,
+      "learning_rate": 4.7488236849061593e-05,
+      "loss": 1.4977,
+      "step": 9512
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8743854761123657,
+      "learning_rate": 4.748797250859107e-05,
+      "loss": 1.407,
+      "step": 9513
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1871321201324463,
+      "learning_rate": 4.748770816812054e-05,
+      "loss": 1.6089,
+      "step": 9514
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8174922466278076,
+      "learning_rate": 4.7487443827650014e-05,
+      "loss": 1.8808,
+      "step": 9515
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.765313744544983,
+      "learning_rate": 4.748717948717949e-05,
+      "loss": 1.3594,
+      "step": 9516
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.009002685546875,
+      "learning_rate": 4.748691514670897e-05,
+      "loss": 1.3899,
+      "step": 9517
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.711877465248108,
+      "learning_rate": 4.7486650806238435e-05,
+      "loss": 1.4627,
+      "step": 9518
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8123483657836914,
+      "learning_rate": 4.7486386465767915e-05,
+      "loss": 1.3525,
+      "step": 9519
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4797672033309937,
+      "learning_rate": 4.748612212529738e-05,
+      "loss": 1.3668,
+      "step": 9520
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.895668625831604,
+      "learning_rate": 4.748585778482686e-05,
+      "loss": 1.5222,
+      "step": 9521
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4464553594589233,
+      "learning_rate": 4.7485593444356336e-05,
+      "loss": 1.1484,
+      "step": 9522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7217395305633545,
+      "learning_rate": 4.748532910388581e-05,
+      "loss": 1.4239,
+      "step": 9523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.121953010559082,
+      "learning_rate": 4.748506476341528e-05,
+      "loss": 1.4065,
+      "step": 9524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0374600887298584,
+      "learning_rate": 4.748480042294475e-05,
+      "loss": 1.4401,
+      "step": 9525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.050203800201416,
+      "learning_rate": 4.748453608247423e-05,
+      "loss": 1.2415,
+      "step": 9526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7158801555633545,
+      "learning_rate": 4.7484271742003704e-05,
+      "loss": 1.4493,
+      "step": 9527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6170676946640015,
+      "learning_rate": 4.748400740153318e-05,
+      "loss": 1.522,
+      "step": 9528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.907333493232727,
+      "learning_rate": 4.748374306106265e-05,
+      "loss": 1.3747,
+      "step": 9529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5675039291381836,
+      "learning_rate": 4.7483478720592125e-05,
+      "loss": 1.356,
+      "step": 9530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2491260766983032,
+      "learning_rate": 4.74832143801216e-05,
+      "loss": 1.353,
+      "step": 9531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.989617943763733,
+      "learning_rate": 4.748295003965108e-05,
+      "loss": 1.3098,
+      "step": 9532
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6041325330734253,
+      "learning_rate": 4.7482685699180545e-05,
+      "loss": 1.5011,
+      "step": 9533
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.230205535888672,
+      "learning_rate": 4.748242135871002e-05,
+      "loss": 1.4092,
+      "step": 9534
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0344135761260986,
+      "learning_rate": 4.748215701823949e-05,
+      "loss": 1.5606,
+      "step": 9535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1623706817626953,
+      "learning_rate": 4.7481892677768966e-05,
+      "loss": 1.6034,
+      "step": 9536
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8207578659057617,
+      "learning_rate": 4.7481628337298446e-05,
+      "loss": 1.3605,
+      "step": 9537
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0558440685272217,
+      "learning_rate": 4.748136399682791e-05,
+      "loss": 1.4637,
+      "step": 9538
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7030664682388306,
+      "learning_rate": 4.7481099656357393e-05,
+      "loss": 1.4142,
+      "step": 9539
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6593883037567139,
+      "learning_rate": 4.748083531588686e-05,
+      "loss": 1.4582,
+      "step": 9540
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4124484062194824,
+      "learning_rate": 4.748057097541634e-05,
+      "loss": 2.1755,
+      "step": 9541
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7539969682693481,
+      "learning_rate": 4.7480306634945814e-05,
+      "loss": 1.3833,
+      "step": 9542
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1499722003936768,
+      "learning_rate": 4.748004229447529e-05,
+      "loss": 1.6929,
+      "step": 9543
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5491374731063843,
+      "learning_rate": 4.747977795400476e-05,
+      "loss": 1.3252,
+      "step": 9544
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5421607494354248,
+      "learning_rate": 4.7479513613534235e-05,
+      "loss": 1.2713,
+      "step": 9545
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3954651355743408,
+      "learning_rate": 4.747924927306371e-05,
+      "loss": 2.1281,
+      "step": 9546
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5606486797332764,
+      "learning_rate": 4.747898493259318e-05,
+      "loss": 1.3677,
+      "step": 9547
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.969041347503662,
+      "learning_rate": 4.7478720592122656e-05,
+      "loss": 1.4666,
+      "step": 9548
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7191671133041382,
+      "learning_rate": 4.747845625165213e-05,
+      "loss": 1.3399,
+      "step": 9549
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.654194951057434,
+      "learning_rate": 4.74781919111816e-05,
+      "loss": 1.3532,
+      "step": 9550
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.881956934928894,
+      "learning_rate": 4.7477927570711076e-05,
+      "loss": 1.513,
+      "step": 9551
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.432395577430725,
+      "learning_rate": 4.747766323024056e-05,
+      "loss": 1.449,
+      "step": 9552
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.627039909362793,
+      "learning_rate": 4.7477398889770023e-05,
+      "loss": 1.2907,
+      "step": 9553
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8012919425964355,
+      "learning_rate": 4.7477134549299504e-05,
+      "loss": 1.4271,
+      "step": 9554
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8238552808761597,
+      "learning_rate": 4.747687020882897e-05,
+      "loss": 1.4664,
+      "step": 9555
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5095633268356323,
+      "learning_rate": 4.747660586835845e-05,
+      "loss": 1.4783,
+      "step": 9556
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.482363224029541,
+      "learning_rate": 4.7476341527887924e-05,
+      "loss": 1.233,
+      "step": 9557
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9418418407440186,
+      "learning_rate": 4.74760771874174e-05,
+      "loss": 1.5772,
+      "step": 9558
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8128269910812378,
+      "learning_rate": 4.747581284694687e-05,
+      "loss": 1.4323,
+      "step": 9559
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5712698698043823,
+      "learning_rate": 4.747554850647634e-05,
+      "loss": 1.319,
+      "step": 9560
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.585814118385315,
+      "learning_rate": 4.747528416600582e-05,
+      "loss": 1.2444,
+      "step": 9561
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5558348894119263,
+      "learning_rate": 4.747501982553529e-05,
+      "loss": 2.1146,
+      "step": 9562
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.989621877670288,
+      "learning_rate": 4.7474755485064766e-05,
+      "loss": 1.5089,
+      "step": 9563
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7085071802139282,
+      "learning_rate": 4.747449114459424e-05,
+      "loss": 1.468,
+      "step": 9564
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.752550482749939,
+      "learning_rate": 4.747422680412371e-05,
+      "loss": 1.4289,
+      "step": 9565
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4283883571624756,
+      "learning_rate": 4.747396246365319e-05,
+      "loss": 1.244,
+      "step": 9566
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7419520616531372,
+      "learning_rate": 4.747369812318267e-05,
+      "loss": 1.4313,
+      "step": 9567
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.718534231185913,
+      "learning_rate": 4.7473433782712134e-05,
+      "loss": 1.1831,
+      "step": 9568
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.893439769744873,
+      "learning_rate": 4.747316944224161e-05,
+      "loss": 1.2983,
+      "step": 9569
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6674944162368774,
+      "learning_rate": 4.747290510177108e-05,
+      "loss": 1.4022,
+      "step": 9570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1095476150512695,
+      "learning_rate": 4.7472640761300554e-05,
+      "loss": 1.5731,
+      "step": 9571
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.923275113105774,
+      "learning_rate": 4.7472376420830035e-05,
+      "loss": 1.5391,
+      "step": 9572
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.022378921508789,
+      "learning_rate": 4.74721120803595e-05,
+      "loss": 1.4409,
+      "step": 9573
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.713544487953186,
+      "learning_rate": 4.747184773988898e-05,
+      "loss": 1.4254,
+      "step": 9574
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6464409828186035,
+      "learning_rate": 4.747158339941845e-05,
+      "loss": 1.2826,
+      "step": 9575
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.660865306854248,
+      "learning_rate": 4.747131905894793e-05,
+      "loss": 1.416,
+      "step": 9576
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.805884599685669,
+      "learning_rate": 4.74710547184774e-05,
+      "loss": 2.2195,
+      "step": 9577
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4019618034362793,
+      "learning_rate": 4.7470790378006876e-05,
+      "loss": 1.4428,
+      "step": 9578
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.337557554244995,
+      "learning_rate": 4.747052603753635e-05,
+      "loss": 1.3364,
+      "step": 9579
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.770188331604004,
+      "learning_rate": 4.7470261697065823e-05,
+      "loss": 1.2961,
+      "step": 9580
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6893242597579956,
+      "learning_rate": 4.74699973565953e-05,
+      "loss": 1.297,
+      "step": 9581
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9240972995758057,
+      "learning_rate": 4.746973301612477e-05,
+      "loss": 1.4013,
+      "step": 9582
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6100890636444092,
+      "learning_rate": 4.7469468675654244e-05,
+      "loss": 1.4243,
+      "step": 9583
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4017595052719116,
+      "learning_rate": 4.746920433518372e-05,
+      "loss": 1.3643,
+      "step": 9584
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1037697792053223,
+      "learning_rate": 4.746893999471319e-05,
+      "loss": 1.4482,
+      "step": 9585
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.203172445297241,
+      "learning_rate": 4.7468675654242665e-05,
+      "loss": 1.299,
+      "step": 9586
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.943371295928955,
+      "learning_rate": 4.7468411313772145e-05,
+      "loss": 1.4209,
+      "step": 9587
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6764118671417236,
+      "learning_rate": 4.746814697330161e-05,
+      "loss": 1.3802,
+      "step": 9588
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.837254762649536,
+      "learning_rate": 4.746788263283109e-05,
+      "loss": 1.4693,
+      "step": 9589
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0184404850006104,
+      "learning_rate": 4.746761829236056e-05,
+      "loss": 1.362,
+      "step": 9590
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6586809158325195,
+      "learning_rate": 4.746735395189004e-05,
+      "loss": 1.2715,
+      "step": 9591
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9838061332702637,
+      "learning_rate": 4.746708961141951e-05,
+      "loss": 1.5012,
+      "step": 9592
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6946728229522705,
+      "learning_rate": 4.7466825270948987e-05,
+      "loss": 1.3647,
+      "step": 9593
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5490919351577759,
+      "learning_rate": 4.746656093047846e-05,
+      "loss": 1.3655,
+      "step": 9594
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0144481658935547,
+      "learning_rate": 4.7466296590007934e-05,
+      "loss": 1.5318,
+      "step": 9595
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5319844484329224,
+      "learning_rate": 4.746603224953741e-05,
+      "loss": 1.2371,
+      "step": 9596
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5843915939331055,
+      "learning_rate": 4.746576790906688e-05,
+      "loss": 1.3256,
+      "step": 9597
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1206421852111816,
+      "learning_rate": 4.7465503568596354e-05,
+      "loss": 1.9031,
+      "step": 9598
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.107126235961914,
+      "learning_rate": 4.746523922812583e-05,
+      "loss": 1.5383,
+      "step": 9599
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9196099042892456,
+      "learning_rate": 4.74649748876553e-05,
+      "loss": 1.4191,
+      "step": 9600
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7550147771835327,
+      "learning_rate": 4.7464710547184775e-05,
+      "loss": 2.6321,
+      "step": 9601
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.653245449066162,
+      "learning_rate": 4.7464446206714256e-05,
+      "loss": 1.4872,
+      "step": 9602
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6866252422332764,
+      "learning_rate": 4.746418186624372e-05,
+      "loss": 1.5744,
+      "step": 9603
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7784332036972046,
+      "learning_rate": 4.74639175257732e-05,
+      "loss": 1.449,
+      "step": 9604
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7498291730880737,
+      "learning_rate": 4.746365318530267e-05,
+      "loss": 1.3524,
+      "step": 9605
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1738686561584473,
+      "learning_rate": 4.746338884483214e-05,
+      "loss": 1.4887,
+      "step": 9606
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.732990026473999,
+      "learning_rate": 4.746312450436162e-05,
+      "loss": 1.3277,
+      "step": 9607
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.955315351486206,
+      "learning_rate": 4.746286016389109e-05,
+      "loss": 1.5432,
+      "step": 9608
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.649092435836792,
+      "learning_rate": 4.746259582342057e-05,
+      "loss": 1.3808,
+      "step": 9609
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.155184030532837,
+      "learning_rate": 4.746233148295004e-05,
+      "loss": 1.6051,
+      "step": 9610
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7125099897384644,
+      "learning_rate": 4.746206714247952e-05,
+      "loss": 1.3764,
+      "step": 9611
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6107704639434814,
+      "learning_rate": 4.746180280200899e-05,
+      "loss": 1.5475,
+      "step": 9612
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4977948665618896,
+      "learning_rate": 4.7461538461538465e-05,
+      "loss": 1.3741,
+      "step": 9613
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.833153247833252,
+      "learning_rate": 4.746127412106794e-05,
+      "loss": 1.408,
+      "step": 9614
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.796120285987854,
+      "learning_rate": 4.746100978059741e-05,
+      "loss": 1.529,
+      "step": 9615
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7640955448150635,
+      "learning_rate": 4.7460745440126885e-05,
+      "loss": 1.4216,
+      "step": 9616
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4285523891448975,
+      "learning_rate": 4.746048109965636e-05,
+      "loss": 1.2152,
+      "step": 9617
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.145545244216919,
+      "learning_rate": 4.746021675918583e-05,
+      "loss": 1.5089,
+      "step": 9618
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.377627372741699,
+      "learning_rate": 4.7459952418715306e-05,
+      "loss": 1.6058,
+      "step": 9619
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.216913938522339,
+      "learning_rate": 4.745968807824478e-05,
+      "loss": 1.529,
+      "step": 9620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.418597936630249,
+      "learning_rate": 4.745942373777425e-05,
+      "loss": 1.288,
+      "step": 9621
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9916510581970215,
+      "learning_rate": 4.7459159397303734e-05,
+      "loss": 1.5904,
+      "step": 9622
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6213277578353882,
+      "learning_rate": 4.74588950568332e-05,
+      "loss": 1.2024,
+      "step": 9623
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0658655166625977,
+      "learning_rate": 4.745863071636268e-05,
+      "loss": 1.451,
+      "step": 9624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.186540126800537,
+      "learning_rate": 4.745836637589215e-05,
+      "loss": 1.4749,
+      "step": 9625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.574039101600647,
+      "learning_rate": 4.745810203542163e-05,
+      "loss": 1.4474,
+      "step": 9626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5100502967834473,
+      "learning_rate": 4.74578376949511e-05,
+      "loss": 1.5489,
+      "step": 9627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.743053436279297,
+      "learning_rate": 4.7457573354480575e-05,
+      "loss": 1.607,
+      "step": 9628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9577529430389404,
+      "learning_rate": 4.745730901401005e-05,
+      "loss": 1.4068,
+      "step": 9629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5306066274642944,
+      "learning_rate": 4.745704467353952e-05,
+      "loss": 1.6286,
+      "step": 9630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.922335147857666,
+      "learning_rate": 4.7456780333068996e-05,
+      "loss": 1.5359,
+      "step": 9631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.551352024078369,
+      "learning_rate": 4.745651599259847e-05,
+      "loss": 1.7318,
+      "step": 9632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.76602041721344,
+      "learning_rate": 4.745625165212794e-05,
+      "loss": 1.467,
+      "step": 9633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5804157257080078,
+      "learning_rate": 4.7455987311657417e-05,
+      "loss": 1.3294,
+      "step": 9634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7439656257629395,
+      "learning_rate": 4.745572297118689e-05,
+      "loss": 1.3414,
+      "step": 9635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9399181604385376,
+      "learning_rate": 4.7455458630716364e-05,
+      "loss": 1.5481,
+      "step": 9636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.613555908203125,
+      "learning_rate": 4.7455194290245844e-05,
+      "loss": 1.4177,
+      "step": 9637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.094956636428833,
+      "learning_rate": 4.745492994977531e-05,
+      "loss": 1.4867,
+      "step": 9638
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.402575135231018,
+      "learning_rate": 4.745466560930479e-05,
+      "loss": 1.2628,
+      "step": 9639
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.289454936981201,
+      "learning_rate": 4.745440126883426e-05,
+      "loss": 1.3891,
+      "step": 9640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7332897186279297,
+      "learning_rate": 4.745413692836373e-05,
+      "loss": 1.4564,
+      "step": 9641
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5808717012405396,
+      "learning_rate": 4.745387258789321e-05,
+      "loss": 1.5167,
+      "step": 9642
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1531455516815186,
+      "learning_rate": 4.745360824742268e-05,
+      "loss": 1.417,
+      "step": 9643
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.676150918006897,
+      "learning_rate": 4.745334390695216e-05,
+      "loss": 1.3907,
+      "step": 9644
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0714051723480225,
+      "learning_rate": 4.7453079566481626e-05,
+      "loss": 1.4503,
+      "step": 9645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0101633071899414,
+      "learning_rate": 4.7452815226011106e-05,
+      "loss": 1.4754,
+      "step": 9646
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6877588033676147,
+      "learning_rate": 4.745255088554058e-05,
+      "loss": 1.3363,
+      "step": 9647
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6640126705169678,
+      "learning_rate": 4.745228654507005e-05,
+      "loss": 1.316,
+      "step": 9648
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4534329175949097,
+      "learning_rate": 4.745202220459953e-05,
+      "loss": 1.4007,
+      "step": 9649
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.546905040740967,
+      "learning_rate": 4.7451757864129e-05,
+      "loss": 1.5078,
+      "step": 9650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4772816896438599,
+      "learning_rate": 4.7451493523658474e-05,
+      "loss": 1.2908,
+      "step": 9651
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7103267908096313,
+      "learning_rate": 4.745122918318795e-05,
+      "loss": 1.4372,
+      "step": 9652
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4647084474563599,
+      "learning_rate": 4.745096484271742e-05,
+      "loss": 1.5205,
+      "step": 9653
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.155710458755493,
+      "learning_rate": 4.7450700502246895e-05,
+      "loss": 1.6283,
+      "step": 9654
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8258103132247925,
+      "learning_rate": 4.745043616177637e-05,
+      "loss": 1.3958,
+      "step": 9655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9847781658172607,
+      "learning_rate": 4.745017182130584e-05,
+      "loss": 1.3902,
+      "step": 9656
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5823466777801514,
+      "learning_rate": 4.744990748083532e-05,
+      "loss": 1.4128,
+      "step": 9657
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0755245685577393,
+      "learning_rate": 4.744964314036479e-05,
+      "loss": 1.5088,
+      "step": 9658
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.07718825340271,
+      "learning_rate": 4.744937879989427e-05,
+      "loss": 1.4517,
+      "step": 9659
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8426368236541748,
+      "learning_rate": 4.7449114459423736e-05,
+      "loss": 1.4361,
+      "step": 9660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6893739700317383,
+      "learning_rate": 4.7448850118953217e-05,
+      "loss": 1.4094,
+      "step": 9661
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6913652420043945,
+      "learning_rate": 4.744858577848269e-05,
+      "loss": 1.3813,
+      "step": 9662
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0016121864318848,
+      "learning_rate": 4.7448321438012164e-05,
+      "loss": 1.3334,
+      "step": 9663
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6106359958648682,
+      "learning_rate": 4.744805709754164e-05,
+      "loss": 1.5041,
+      "step": 9664
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.985791802406311,
+      "learning_rate": 4.744779275707111e-05,
+      "loss": 1.5653,
+      "step": 9665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9417072534561157,
+      "learning_rate": 4.7447528416600584e-05,
+      "loss": 1.3529,
+      "step": 9666
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9011170864105225,
+      "learning_rate": 4.744726407613006e-05,
+      "loss": 1.4404,
+      "step": 9667
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1095898151397705,
+      "learning_rate": 4.744699973565953e-05,
+      "loss": 1.4228,
+      "step": 9668
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.546808123588562,
+      "learning_rate": 4.7446735395189005e-05,
+      "loss": 1.4263,
+      "step": 9669
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6418075561523438,
+      "learning_rate": 4.744647105471848e-05,
+      "loss": 1.4353,
+      "step": 9670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7156964540481567,
+      "learning_rate": 4.744620671424795e-05,
+      "loss": 1.4833,
+      "step": 9671
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4025241136550903,
+      "learning_rate": 4.744594237377743e-05,
+      "loss": 1.3307,
+      "step": 9672
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6571849584579468,
+      "learning_rate": 4.74456780333069e-05,
+      "loss": 1.6198,
+      "step": 9673
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.579200267791748,
+      "learning_rate": 4.744541369283638e-05,
+      "loss": 2.675,
+      "step": 9674
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7741514444351196,
+      "learning_rate": 4.7445149352365847e-05,
+      "loss": 1.4033,
+      "step": 9675
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6565134525299072,
+      "learning_rate": 4.744488501189533e-05,
+      "loss": 1.3987,
+      "step": 9676
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5531623363494873,
+      "learning_rate": 4.74446206714248e-05,
+      "loss": 1.1641,
+      "step": 9677
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2336833477020264,
+      "learning_rate": 4.744435633095427e-05,
+      "loss": 2.4365,
+      "step": 9678
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5807360410690308,
+      "learning_rate": 4.744409199048375e-05,
+      "loss": 1.3711,
+      "step": 9679
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.731528401374817,
+      "learning_rate": 4.7443827650013214e-05,
+      "loss": 1.4445,
+      "step": 9680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.814948558807373,
+      "learning_rate": 4.7443563309542695e-05,
+      "loss": 1.5035,
+      "step": 9681
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.595733642578125,
+      "learning_rate": 4.744329896907217e-05,
+      "loss": 1.4564,
+      "step": 9682
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9052565097808838,
+      "learning_rate": 4.744303462860164e-05,
+      "loss": 1.3772,
+      "step": 9683
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.755692481994629,
+      "learning_rate": 4.7442770288131115e-05,
+      "loss": 1.3452,
+      "step": 9684
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8075692653656006,
+      "learning_rate": 4.744250594766059e-05,
+      "loss": 1.6422,
+      "step": 9685
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.143083095550537,
+      "learning_rate": 4.744224160719006e-05,
+      "loss": 1.3937,
+      "step": 9686
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0616016387939453,
+      "learning_rate": 4.7441977266719536e-05,
+      "loss": 1.5039,
+      "step": 9687
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.905427098274231,
+      "learning_rate": 4.744171292624901e-05,
+      "loss": 1.463,
+      "step": 9688
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.188793420791626,
+      "learning_rate": 4.744144858577848e-05,
+      "loss": 1.5336,
+      "step": 9689
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.354365110397339,
+      "learning_rate": 4.744118424530796e-05,
+      "loss": 1.466,
+      "step": 9690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8842806816101074,
+      "learning_rate": 4.744091990483743e-05,
+      "loss": 1.4149,
+      "step": 9691
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.657085418701172,
+      "learning_rate": 4.744065556436691e-05,
+      "loss": 1.5142,
+      "step": 9692
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6826249361038208,
+      "learning_rate": 4.744039122389638e-05,
+      "loss": 1.455,
+      "step": 9693
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6726666688919067,
+      "learning_rate": 4.744012688342586e-05,
+      "loss": 1.2611,
+      "step": 9694
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0013644695281982,
+      "learning_rate": 4.7439862542955325e-05,
+      "loss": 2.4732,
+      "step": 9695
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.52456533908844,
+      "learning_rate": 4.7439598202484805e-05,
+      "loss": 1.3371,
+      "step": 9696
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9760278463363647,
+      "learning_rate": 4.743933386201428e-05,
+      "loss": 1.4576,
+      "step": 9697
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.245453119277954,
+      "learning_rate": 4.743906952154375e-05,
+      "loss": 1.5592,
+      "step": 9698
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4043936729431152,
+      "learning_rate": 4.7438805181073226e-05,
+      "loss": 2.1634,
+      "step": 9699
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6834800243377686,
+      "learning_rate": 4.74385408406027e-05,
+      "loss": 1.3872,
+      "step": 9700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6424251794815063,
+      "learning_rate": 4.743827650013217e-05,
+      "loss": 1.3711,
+      "step": 9701
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7000664472579956,
+      "learning_rate": 4.7438012159661646e-05,
+      "loss": 1.4309,
+      "step": 9702
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3224856853485107,
+      "learning_rate": 4.743774781919112e-05,
+      "loss": 1.5475,
+      "step": 9703
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7742178440093994,
+      "learning_rate": 4.7437483478720594e-05,
+      "loss": 1.4604,
+      "step": 9704
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.676611304283142,
+      "learning_rate": 4.743721913825007e-05,
+      "loss": 1.3798,
+      "step": 9705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6836692094802856,
+      "learning_rate": 4.743695479777954e-05,
+      "loss": 1.4746,
+      "step": 9706
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2975218296051025,
+      "learning_rate": 4.743669045730902e-05,
+      "loss": 1.1595,
+      "step": 9707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0111777782440186,
+      "learning_rate": 4.743642611683849e-05,
+      "loss": 1.3714,
+      "step": 9708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6901402473449707,
+      "learning_rate": 4.743616177636797e-05,
+      "loss": 1.3256,
+      "step": 9709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.06223726272583,
+      "learning_rate": 4.7435897435897435e-05,
+      "loss": 1.5137,
+      "step": 9710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8506999015808105,
+      "learning_rate": 4.7435633095426915e-05,
+      "loss": 1.4013,
+      "step": 9711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3629817962646484,
+      "learning_rate": 4.743536875495639e-05,
+      "loss": 1.4803,
+      "step": 9712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.869387149810791,
+      "learning_rate": 4.743510441448586e-05,
+      "loss": 1.3391,
+      "step": 9713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.62717866897583,
+      "learning_rate": 4.7434840074015336e-05,
+      "loss": 1.3294,
+      "step": 9714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8911221027374268,
+      "learning_rate": 4.74345757335448e-05,
+      "loss": 1.4392,
+      "step": 9715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4218206405639648,
+      "learning_rate": 4.743431139307428e-05,
+      "loss": 1.1529,
+      "step": 9716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7956126928329468,
+      "learning_rate": 4.743404705260376e-05,
+      "loss": 1.3787,
+      "step": 9717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.847095012664795,
+      "learning_rate": 4.743378271213323e-05,
+      "loss": 1.4003,
+      "step": 9718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9821817874908447,
+      "learning_rate": 4.7433518371662704e-05,
+      "loss": 1.542,
+      "step": 9719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.121551513671875,
+      "learning_rate": 4.743325403119218e-05,
+      "loss": 1.5187,
+      "step": 9720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.446273922920227,
+      "learning_rate": 4.743298969072165e-05,
+      "loss": 1.2983,
+      "step": 9721
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.874976634979248,
+      "learning_rate": 4.743272535025113e-05,
+      "loss": 1.4431,
+      "step": 9722
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.582619309425354,
+      "learning_rate": 4.74324610097806e-05,
+      "loss": 1.2916,
+      "step": 9723
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.089562177658081,
+      "learning_rate": 4.743219666931007e-05,
+      "loss": 1.4016,
+      "step": 9724
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6443352699279785,
+      "learning_rate": 4.7431932328839545e-05,
+      "loss": 1.1729,
+      "step": 9725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8028582334518433,
+      "learning_rate": 4.743166798836902e-05,
+      "loss": 1.3897,
+      "step": 9726
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4835513830184937,
+      "learning_rate": 4.74314036478985e-05,
+      "loss": 1.1926,
+      "step": 9727
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8520070314407349,
+      "learning_rate": 4.7431139307427966e-05,
+      "loss": 1.4614,
+      "step": 9728
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2265214920043945,
+      "learning_rate": 4.7430874966957446e-05,
+      "loss": 1.6833,
+      "step": 9729
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6370835304260254,
+      "learning_rate": 4.743061062648691e-05,
+      "loss": 1.4406,
+      "step": 9730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4863029718399048,
+      "learning_rate": 4.7430346286016394e-05,
+      "loss": 1.3916,
+      "step": 9731
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.278444290161133,
+      "learning_rate": 4.743008194554587e-05,
+      "loss": 1.4211,
+      "step": 9732
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4264566898345947,
+      "learning_rate": 4.742981760507534e-05,
+      "loss": 1.6576,
+      "step": 9733
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.538214921951294,
+      "learning_rate": 4.7429553264604814e-05,
+      "loss": 1.5424,
+      "step": 9734
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5793455839157104,
+      "learning_rate": 4.742928892413429e-05,
+      "loss": 1.3807,
+      "step": 9735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.011232852935791,
+      "learning_rate": 4.742902458366376e-05,
+      "loss": 1.4425,
+      "step": 9736
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5147942304611206,
+      "learning_rate": 4.7428760243193235e-05,
+      "loss": 1.3512,
+      "step": 9737
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8037478923797607,
+      "learning_rate": 4.742849590272271e-05,
+      "loss": 1.4216,
+      "step": 9738
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6654729843139648,
+      "learning_rate": 4.742823156225218e-05,
+      "loss": 1.3842,
+      "step": 9739
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6565021276474,
+      "learning_rate": 4.7427967221781656e-05,
+      "loss": 1.427,
+      "step": 9740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6219584941864014,
+      "learning_rate": 4.742770288131113e-05,
+      "loss": 1.4374,
+      "step": 9741
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9064425230026245,
+      "learning_rate": 4.742743854084061e-05,
+      "loss": 1.3697,
+      "step": 9742
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7287344932556152,
+      "learning_rate": 4.7427174200370076e-05,
+      "loss": 1.319,
+      "step": 9743
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8740429878234863,
+      "learning_rate": 4.742690985989956e-05,
+      "loss": 1.4307,
+      "step": 9744
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1346962451934814,
+      "learning_rate": 4.7426645519429024e-05,
+      "loss": 1.5167,
+      "step": 9745
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.074317693710327,
+      "learning_rate": 4.7426381178958504e-05,
+      "loss": 1.261,
+      "step": 9746
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.715406060218811,
+      "learning_rate": 4.742611683848798e-05,
+      "loss": 1.4873,
+      "step": 9747
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.392186164855957,
+      "learning_rate": 4.742585249801745e-05,
+      "loss": 1.4042,
+      "step": 9748
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3644341230392456,
+      "learning_rate": 4.7425588157546925e-05,
+      "loss": 1.3247,
+      "step": 9749
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7654374837875366,
+      "learning_rate": 4.742532381707639e-05,
+      "loss": 1.42,
+      "step": 9750
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.095043182373047,
+      "learning_rate": 4.742505947660587e-05,
+      "loss": 1.379,
+      "step": 9751
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6926509141921997,
+      "learning_rate": 4.7424795136135345e-05,
+      "loss": 1.4866,
+      "step": 9752
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.452970266342163,
+      "learning_rate": 4.742453079566482e-05,
+      "loss": 1.3819,
+      "step": 9753
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9122223854064941,
+      "learning_rate": 4.742426645519429e-05,
+      "loss": 1.4051,
+      "step": 9754
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.062351942062378,
+      "learning_rate": 4.7424002114723766e-05,
+      "loss": 1.2946,
+      "step": 9755
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.99470055103302,
+      "learning_rate": 4.742373777425324e-05,
+      "loss": 1.5276,
+      "step": 9756
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.93808114528656,
+      "learning_rate": 4.742347343378272e-05,
+      "loss": 1.3697,
+      "step": 9757
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1159136295318604,
+      "learning_rate": 4.742320909331219e-05,
+      "loss": 1.5858,
+      "step": 9758
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5921502113342285,
+      "learning_rate": 4.742294475284166e-05,
+      "loss": 1.4019,
+      "step": 9759
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2721312046051025,
+      "learning_rate": 4.7422680412371134e-05,
+      "loss": 1.5219,
+      "step": 9760
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6189414262771606,
+      "learning_rate": 4.742241607190061e-05,
+      "loss": 1.3114,
+      "step": 9761
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.232961654663086,
+      "learning_rate": 4.742215173143009e-05,
+      "loss": 1.4621,
+      "step": 9762
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7617441415786743,
+      "learning_rate": 4.7421887390959555e-05,
+      "loss": 1.651,
+      "step": 9763
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7587215900421143,
+      "learning_rate": 4.7421623050489035e-05,
+      "loss": 1.1511,
+      "step": 9764
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8673911094665527,
+      "learning_rate": 4.74213587100185e-05,
+      "loss": 1.5226,
+      "step": 9765
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.154116153717041,
+      "learning_rate": 4.742109436954798e-05,
+      "loss": 1.4324,
+      "step": 9766
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2623462677001953,
+      "learning_rate": 4.7420830029077456e-05,
+      "loss": 1.2746,
+      "step": 9767
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.74824857711792,
+      "learning_rate": 4.742056568860693e-05,
+      "loss": 1.3155,
+      "step": 9768
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6098318099975586,
+      "learning_rate": 4.74203013481364e-05,
+      "loss": 1.4392,
+      "step": 9769
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4757299423217773,
+      "learning_rate": 4.7420037007665876e-05,
+      "loss": 1.2483,
+      "step": 9770
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6621023416519165,
+      "learning_rate": 4.741977266719535e-05,
+      "loss": 1.223,
+      "step": 9771
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6494404077529907,
+      "learning_rate": 4.7419508326724824e-05,
+      "loss": 1.3869,
+      "step": 9772
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.341463565826416,
+      "learning_rate": 4.74192439862543e-05,
+      "loss": 1.4168,
+      "step": 9773
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6868599653244019,
+      "learning_rate": 4.741897964578377e-05,
+      "loss": 1.3006,
+      "step": 9774
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5780116319656372,
+      "learning_rate": 4.7418715305313244e-05,
+      "loss": 1.3765,
+      "step": 9775
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.622246265411377,
+      "learning_rate": 4.741845096484272e-05,
+      "loss": 1.3579,
+      "step": 9776
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6105676889419556,
+      "learning_rate": 4.74181866243722e-05,
+      "loss": 1.3309,
+      "step": 9777
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6623133420944214,
+      "learning_rate": 4.7417922283901665e-05,
+      "loss": 1.468,
+      "step": 9778
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8102198839187622,
+      "learning_rate": 4.7417657943431145e-05,
+      "loss": 1.4689,
+      "step": 9779
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6464463472366333,
+      "learning_rate": 4.741739360296061e-05,
+      "loss": 1.2437,
+      "step": 9780
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.306691884994507,
+      "learning_rate": 4.741712926249009e-05,
+      "loss": 1.3369,
+      "step": 9781
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3445322513580322,
+      "learning_rate": 4.7416864922019566e-05,
+      "loss": 1.6031,
+      "step": 9782
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9786163568496704,
+      "learning_rate": 4.741660058154904e-05,
+      "loss": 1.4122,
+      "step": 9783
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8590017557144165,
+      "learning_rate": 4.741633624107851e-05,
+      "loss": 1.353,
+      "step": 9784
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6323165893554688,
+      "learning_rate": 4.741607190060799e-05,
+      "loss": 1.3319,
+      "step": 9785
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6807843446731567,
+      "learning_rate": 4.741580756013746e-05,
+      "loss": 1.4169,
+      "step": 9786
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5634254217147827,
+      "learning_rate": 4.7415543219666934e-05,
+      "loss": 1.396,
+      "step": 9787
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0057811737060547,
+      "learning_rate": 4.741527887919641e-05,
+      "loss": 1.432,
+      "step": 9788
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7403788566589355,
+      "learning_rate": 4.741501453872588e-05,
+      "loss": 1.4129,
+      "step": 9789
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6924303770065308,
+      "learning_rate": 4.7414750198255355e-05,
+      "loss": 1.4666,
+      "step": 9790
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5304319858551025,
+      "learning_rate": 4.741448585778483e-05,
+      "loss": 1.4513,
+      "step": 9791
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6806623935699463,
+      "learning_rate": 4.741422151731431e-05,
+      "loss": 1.3446,
+      "step": 9792
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.421452045440674,
+      "learning_rate": 4.7413957176843775e-05,
+      "loss": 1.416,
+      "step": 9793
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2778563499450684,
+      "learning_rate": 4.7413692836373256e-05,
+      "loss": 1.4938,
+      "step": 9794
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7417933940887451,
+      "learning_rate": 4.741342849590272e-05,
+      "loss": 1.3988,
+      "step": 9795
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2458492517471313,
+      "learning_rate": 4.7413164155432196e-05,
+      "loss": 1.2046,
+      "step": 9796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6505032777786255,
+      "learning_rate": 4.7412899814961676e-05,
+      "loss": 1.3647,
+      "step": 9797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6171071529388428,
+      "learning_rate": 4.741263547449114e-05,
+      "loss": 1.3234,
+      "step": 9798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3907227516174316,
+      "learning_rate": 4.7412371134020624e-05,
+      "loss": 1.7105,
+      "step": 9799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7313731908798218,
+      "learning_rate": 4.741210679355009e-05,
+      "loss": 1.4253,
+      "step": 9800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7746590375900269,
+      "learning_rate": 4.741184245307957e-05,
+      "loss": 1.4556,
+      "step": 9801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.365588426589966,
+      "learning_rate": 4.7411578112609044e-05,
+      "loss": 1.5461,
+      "step": 9802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0022037029266357,
+      "learning_rate": 4.741131377213852e-05,
+      "loss": 1.3606,
+      "step": 9803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7282719612121582,
+      "learning_rate": 4.741104943166799e-05,
+      "loss": 1.2251,
+      "step": 9804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7760536670684814,
+      "learning_rate": 4.7410785091197465e-05,
+      "loss": 1.2301,
+      "step": 9805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6473822593688965,
+      "learning_rate": 4.741052075072694e-05,
+      "loss": 1.3414,
+      "step": 9806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.406525135040283,
+      "learning_rate": 4.741025641025641e-05,
+      "loss": 2.7799,
+      "step": 9807
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5717401504516602,
+      "learning_rate": 4.7409992069785886e-05,
+      "loss": 1.3709,
+      "step": 9808
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7900104522705078,
+      "learning_rate": 4.740972772931536e-05,
+      "loss": 1.477,
+      "step": 9809
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6115936040878296,
+      "learning_rate": 4.740946338884483e-05,
+      "loss": 1.317,
+      "step": 9810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5318025350570679,
+      "learning_rate": 4.7409199048374306e-05,
+      "loss": 1.3947,
+      "step": 9811
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6926288604736328,
+      "learning_rate": 4.740893470790379e-05,
+      "loss": 1.4222,
+      "step": 9812
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7887133359909058,
+      "learning_rate": 4.7408670367433254e-05,
+      "loss": 1.6181,
+      "step": 9813
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.637712001800537,
+      "learning_rate": 4.7408406026962734e-05,
+      "loss": 1.3178,
+      "step": 9814
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9834357500076294,
+      "learning_rate": 4.74081416864922e-05,
+      "loss": 1.5158,
+      "step": 9815
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.992006778717041,
+      "learning_rate": 4.740787734602168e-05,
+      "loss": 1.1887,
+      "step": 9816
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5293161869049072,
+      "learning_rate": 4.7407613005551155e-05,
+      "loss": 1.4738,
+      "step": 9817
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7695237398147583,
+      "learning_rate": 4.740734866508063e-05,
+      "loss": 1.3393,
+      "step": 9818
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2396788597106934,
+      "learning_rate": 4.74070843246101e-05,
+      "loss": 1.5395,
+      "step": 9819
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.007636547088623,
+      "learning_rate": 4.7406819984139575e-05,
+      "loss": 1.3504,
+      "step": 9820
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4509855508804321,
+      "learning_rate": 4.740655564366905e-05,
+      "loss": 1.418,
+      "step": 9821
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.594650149345398,
+      "learning_rate": 4.740629130319852e-05,
+      "loss": 1.4077,
+      "step": 9822
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.624462366104126,
+      "learning_rate": 4.7406026962727996e-05,
+      "loss": 1.4578,
+      "step": 9823
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9085613489151,
+      "learning_rate": 4.740576262225747e-05,
+      "loss": 1.351,
+      "step": 9824
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7779464721679688,
+      "learning_rate": 4.740549828178694e-05,
+      "loss": 1.4124,
+      "step": 9825
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.540848731994629,
+      "learning_rate": 4.740523394131642e-05,
+      "loss": 1.238,
+      "step": 9826
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4889315366744995,
+      "learning_rate": 4.74049696008459e-05,
+      "loss": 1.2558,
+      "step": 9827
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3165173530578613,
+      "learning_rate": 4.7404705260375364e-05,
+      "loss": 1.5604,
+      "step": 9828
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7384892702102661,
+      "learning_rate": 4.7404440919904844e-05,
+      "loss": 1.3408,
+      "step": 9829
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7127107381820679,
+      "learning_rate": 4.740417657943431e-05,
+      "loss": 1.529,
+      "step": 9830
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5684272050857544,
+      "learning_rate": 4.740391223896379e-05,
+      "loss": 1.3988,
+      "step": 9831
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8438987731933594,
+      "learning_rate": 4.7403647898493265e-05,
+      "loss": 1.4083,
+      "step": 9832
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.599011182785034,
+      "learning_rate": 4.740338355802273e-05,
+      "loss": 1.6139,
+      "step": 9833
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.217912197113037,
+      "learning_rate": 4.740311921755221e-05,
+      "loss": 1.5454,
+      "step": 9834
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6630778312683105,
+      "learning_rate": 4.740285487708168e-05,
+      "loss": 1.3355,
+      "step": 9835
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4266283512115479,
+      "learning_rate": 4.740259053661116e-05,
+      "loss": 1.2566,
+      "step": 9836
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.453277111053467,
+      "learning_rate": 4.740232619614063e-05,
+      "loss": 1.4721,
+      "step": 9837
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5035127401351929,
+      "learning_rate": 4.7402061855670106e-05,
+      "loss": 1.3327,
+      "step": 9838
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.617296576499939,
+      "learning_rate": 4.740179751519958e-05,
+      "loss": 1.2877,
+      "step": 9839
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.178837776184082,
+      "learning_rate": 4.7401533174729053e-05,
+      "loss": 1.5966,
+      "step": 9840
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8683587312698364,
+      "learning_rate": 4.740126883425853e-05,
+      "loss": 1.4792,
+      "step": 9841
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1330952644348145,
+      "learning_rate": 4.7401004493788e-05,
+      "loss": 1.297,
+      "step": 9842
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6719664335250854,
+      "learning_rate": 4.7400740153317474e-05,
+      "loss": 1.3993,
+      "step": 9843
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7444361448287964,
+      "learning_rate": 4.740047581284695e-05,
+      "loss": 1.3411,
+      "step": 9844
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6480286121368408,
+      "learning_rate": 4.740021147237642e-05,
+      "loss": 2.4773,
+      "step": 9845
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.652847409248352,
+      "learning_rate": 4.7399947131905895e-05,
+      "loss": 1.4303,
+      "step": 9846
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4624269008636475,
+      "learning_rate": 4.7399682791435375e-05,
+      "loss": 1.3196,
+      "step": 9847
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.346768856048584,
+      "learning_rate": 4.739941845096484e-05,
+      "loss": 1.3984,
+      "step": 9848
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.000136613845825,
+      "learning_rate": 4.739915411049432e-05,
+      "loss": 1.4947,
+      "step": 9849
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6781648397445679,
+      "learning_rate": 4.739888977002379e-05,
+      "loss": 1.3819,
+      "step": 9850
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4874688386917114,
+      "learning_rate": 4.739862542955327e-05,
+      "loss": 1.2504,
+      "step": 9851
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8763456344604492,
+      "learning_rate": 4.739836108908274e-05,
+      "loss": 2.0907,
+      "step": 9852
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5556373596191406,
+      "learning_rate": 4.739809674861222e-05,
+      "loss": 1.2542,
+      "step": 9853
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7960031032562256,
+      "learning_rate": 4.739783240814169e-05,
+      "loss": 1.3568,
+      "step": 9854
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0331871509552,
+      "learning_rate": 4.7397568067671164e-05,
+      "loss": 1.437,
+      "step": 9855
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7484885454177856,
+      "learning_rate": 4.739730372720064e-05,
+      "loss": 1.2564,
+      "step": 9856
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4289435148239136,
+      "learning_rate": 4.739703938673011e-05,
+      "loss": 1.5449,
+      "step": 9857
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.505388855934143,
+      "learning_rate": 4.7396775046259585e-05,
+      "loss": 1.3624,
+      "step": 9858
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0027096271514893,
+      "learning_rate": 4.739651070578906e-05,
+      "loss": 1.6905,
+      "step": 9859
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1019070148468018,
+      "learning_rate": 4.739624636531853e-05,
+      "loss": 1.4754,
+      "step": 9860
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9022917747497559,
+      "learning_rate": 4.7395982024848005e-05,
+      "loss": 1.3068,
+      "step": 9861
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.586722493171692,
+      "learning_rate": 4.7395717684377486e-05,
+      "loss": 1.4466,
+      "step": 9862
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5714070796966553,
+      "learning_rate": 4.739545334390695e-05,
+      "loss": 1.4085,
+      "step": 9863
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7451313734054565,
+      "learning_rate": 4.739518900343643e-05,
+      "loss": 1.2475,
+      "step": 9864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0968379974365234,
+      "learning_rate": 4.73949246629659e-05,
+      "loss": 1.3233,
+      "step": 9865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.70927095413208,
+      "learning_rate": 4.739466032249538e-05,
+      "loss": 1.3739,
+      "step": 9866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9402939081192017,
+      "learning_rate": 4.7394395982024853e-05,
+      "loss": 1.4849,
+      "step": 9867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.039172887802124,
+      "learning_rate": 4.739413164155432e-05,
+      "loss": 1.3425,
+      "step": 9868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9514241218566895,
+      "learning_rate": 4.73938673010838e-05,
+      "loss": 1.6044,
+      "step": 9869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8596962690353394,
+      "learning_rate": 4.739360296061327e-05,
+      "loss": 1.5246,
+      "step": 9870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.905182123184204,
+      "learning_rate": 4.739333862014275e-05,
+      "loss": 1.4089,
+      "step": 9871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4154140949249268,
+      "learning_rate": 4.739307427967222e-05,
+      "loss": 1.2627,
+      "step": 9872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7239737510681152,
+      "learning_rate": 4.7392809939201695e-05,
+      "loss": 1.3802,
+      "step": 9873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.309044599533081,
+      "learning_rate": 4.739254559873117e-05,
+      "loss": 1.6249,
+      "step": 9874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7331722974777222,
+      "learning_rate": 4.739228125826064e-05,
+      "loss": 1.2772,
+      "step": 9875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7868859767913818,
+      "learning_rate": 4.7392016917790116e-05,
+      "loss": 1.3839,
+      "step": 9876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6242107152938843,
+      "learning_rate": 4.739175257731959e-05,
+      "loss": 1.2616,
+      "step": 9877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6232455968856812,
+      "learning_rate": 4.739148823684906e-05,
+      "loss": 1.4119,
+      "step": 9878
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8667126893997192,
+      "learning_rate": 4.7391223896378536e-05,
+      "loss": 1.497,
+      "step": 9879
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8645045757293701,
+      "learning_rate": 4.739095955590801e-05,
+      "loss": 1.4324,
+      "step": 9880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2314352989196777,
+      "learning_rate": 4.7390695215437483e-05,
+      "loss": 1.5121,
+      "step": 9881
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4330971240997314,
+      "learning_rate": 4.7390430874966964e-05,
+      "loss": 1.2846,
+      "step": 9882
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.807616949081421,
+      "learning_rate": 4.739016653449643e-05,
+      "loss": 1.3685,
+      "step": 9883
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4571599960327148,
+      "learning_rate": 4.738990219402591e-05,
+      "loss": 1.2089,
+      "step": 9884
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.570339322090149,
+      "learning_rate": 4.738963785355538e-05,
+      "loss": 1.2494,
+      "step": 9885
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7612537145614624,
+      "learning_rate": 4.738937351308486e-05,
+      "loss": 1.2862,
+      "step": 9886
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5759142637252808,
+      "learning_rate": 4.738910917261433e-05,
+      "loss": 1.4343,
+      "step": 9887
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.18227219581604,
+      "learning_rate": 4.7388844832143805e-05,
+      "loss": 1.5344,
+      "step": 9888
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3573851585388184,
+      "learning_rate": 4.738858049167328e-05,
+      "loss": 1.9647,
+      "step": 9889
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.134456157684326,
+      "learning_rate": 4.738831615120275e-05,
+      "loss": 1.4442,
+      "step": 9890
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1724445819854736,
+      "learning_rate": 4.7388051810732226e-05,
+      "loss": 1.5238,
+      "step": 9891
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8633109331130981,
+      "learning_rate": 4.73877874702617e-05,
+      "loss": 1.5023,
+      "step": 9892
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7727493047714233,
+      "learning_rate": 4.738752312979117e-05,
+      "loss": 1.348,
+      "step": 9893
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2598767280578613,
+      "learning_rate": 4.738725878932065e-05,
+      "loss": 1.5319,
+      "step": 9894
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6037031412124634,
+      "learning_rate": 4.738699444885012e-05,
+      "loss": 1.3526,
+      "step": 9895
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8117445707321167,
+      "learning_rate": 4.7386730108379594e-05,
+      "loss": 1.4878,
+      "step": 9896
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.291402816772461,
+      "learning_rate": 4.7386465767909074e-05,
+      "loss": 1.3328,
+      "step": 9897
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8151530027389526,
+      "learning_rate": 4.738620142743854e-05,
+      "loss": 1.4231,
+      "step": 9898
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.852590799331665,
+      "learning_rate": 4.738593708696802e-05,
+      "loss": 1.4605,
+      "step": 9899
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7571502923965454,
+      "learning_rate": 4.738567274649749e-05,
+      "loss": 1.4687,
+      "step": 9900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8995798826217651,
+      "learning_rate": 4.738540840602697e-05,
+      "loss": 1.4075,
+      "step": 9901
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1069862842559814,
+      "learning_rate": 4.738514406555644e-05,
+      "loss": 1.4091,
+      "step": 9902
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.536231517791748,
+      "learning_rate": 4.7384879725085916e-05,
+      "loss": 1.3834,
+      "step": 9903
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6486103534698486,
+      "learning_rate": 4.738461538461539e-05,
+      "loss": 1.4325,
+      "step": 9904
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7112538814544678,
+      "learning_rate": 4.7384351044144856e-05,
+      "loss": 1.339,
+      "step": 9905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2633419036865234,
+      "learning_rate": 4.7384086703674336e-05,
+      "loss": 1.6736,
+      "step": 9906
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0896246433258057,
+      "learning_rate": 4.738382236320381e-05,
+      "loss": 1.4558,
+      "step": 9907
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.063023328781128,
+      "learning_rate": 4.7383558022733283e-05,
+      "loss": 1.2427,
+      "step": 9908
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7697546482086182,
+      "learning_rate": 4.738329368226276e-05,
+      "loss": 1.4483,
+      "step": 9909
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4047762155532837,
+      "learning_rate": 4.738302934179223e-05,
+      "loss": 1.2656,
+      "step": 9910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9402869939804077,
+      "learning_rate": 4.7382765001321704e-05,
+      "loss": 1.3313,
+      "step": 9911
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7829662561416626,
+      "learning_rate": 4.7382500660851184e-05,
+      "loss": 1.4165,
+      "step": 9912
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8303873538970947,
+      "learning_rate": 4.738223632038065e-05,
+      "loss": 1.4967,
+      "step": 9913
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.148130416870117,
+      "learning_rate": 4.7381971979910125e-05,
+      "loss": 1.4044,
+      "step": 9914
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.44137442111969,
+      "learning_rate": 4.73817076394396e-05,
+      "loss": 1.2657,
+      "step": 9915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.340672016143799,
+      "learning_rate": 4.738144329896907e-05,
+      "loss": 1.7037,
+      "step": 9916
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8906359672546387,
+      "learning_rate": 4.738117895849855e-05,
+      "loss": 1.4106,
+      "step": 9917
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7105541229248047,
+      "learning_rate": 4.738091461802802e-05,
+      "loss": 1.263,
+      "step": 9918
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.59499192237854,
+      "learning_rate": 4.73806502775575e-05,
+      "loss": 1.3154,
+      "step": 9919
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.652981996536255,
+      "learning_rate": 4.7380385937086966e-05,
+      "loss": 1.5037,
+      "step": 9920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7226498126983643,
+      "learning_rate": 4.7380121596616447e-05,
+      "loss": 1.2693,
+      "step": 9921
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7463129758834839,
+      "learning_rate": 4.737985725614592e-05,
+      "loss": 1.3646,
+      "step": 9922
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5490920543670654,
+      "learning_rate": 4.7379592915675394e-05,
+      "loss": 1.3086,
+      "step": 9923
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6093355417251587,
+      "learning_rate": 4.737932857520487e-05,
+      "loss": 1.3638,
+      "step": 9924
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7164510488510132,
+      "learning_rate": 4.737906423473434e-05,
+      "loss": 1.3948,
+      "step": 9925
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4024734497070312,
+      "learning_rate": 4.7378799894263814e-05,
+      "loss": 1.6185,
+      "step": 9926
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6083414554595947,
+      "learning_rate": 4.737853555379329e-05,
+      "loss": 1.3721,
+      "step": 9927
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4463368654251099,
+      "learning_rate": 4.737827121332276e-05,
+      "loss": 1.3421,
+      "step": 9928
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2408483028411865,
+      "learning_rate": 4.7378006872852235e-05,
+      "loss": 1.5,
+      "step": 9929
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4324724674224854,
+      "learning_rate": 4.737774253238171e-05,
+      "loss": 1.5284,
+      "step": 9930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9067269563674927,
+      "learning_rate": 4.737747819191118e-05,
+      "loss": 1.4378,
+      "step": 9931
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1581759452819824,
+      "learning_rate": 4.737721385144066e-05,
+      "loss": 1.4582,
+      "step": 9932
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4971866607666016,
+      "learning_rate": 4.737694951097013e-05,
+      "loss": 1.3058,
+      "step": 9933
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8757864236831665,
+      "learning_rate": 4.737668517049961e-05,
+      "loss": 1.378,
+      "step": 9934
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.50247323513031,
+      "learning_rate": 4.7376420830029077e-05,
+      "loss": 1.3583,
+      "step": 9935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1143007278442383,
+      "learning_rate": 4.737615648955856e-05,
+      "loss": 1.6097,
+      "step": 9936
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9796298742294312,
+      "learning_rate": 4.737589214908803e-05,
+      "loss": 1.4953,
+      "step": 9937
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5393462181091309,
+      "learning_rate": 4.7375627808617504e-05,
+      "loss": 1.3209,
+      "step": 9938
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3860467672348022,
+      "learning_rate": 4.737536346814698e-05,
+      "loss": 1.2512,
+      "step": 9939
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.592223048210144,
+      "learning_rate": 4.7375099127676444e-05,
+      "loss": 1.3346,
+      "step": 9940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.293235778808594,
+      "learning_rate": 4.7374834787205925e-05,
+      "loss": 1.5247,
+      "step": 9941
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6316101551055908,
+      "learning_rate": 4.73745704467354e-05,
+      "loss": 1.4557,
+      "step": 9942
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8155215978622437,
+      "learning_rate": 4.737430610626487e-05,
+      "loss": 1.6405,
+      "step": 9943
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8734278678894043,
+      "learning_rate": 4.7374041765794346e-05,
+      "loss": 1.3411,
+      "step": 9944
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8543164730072021,
+      "learning_rate": 4.737377742532382e-05,
+      "loss": 1.4967,
+      "step": 9945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4660303592681885,
+      "learning_rate": 4.737351308485329e-05,
+      "loss": 1.5679,
+      "step": 9946
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8811089992523193,
+      "learning_rate": 4.737324874438277e-05,
+      "loss": 1.6579,
+      "step": 9947
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5360138416290283,
+      "learning_rate": 4.737298440391224e-05,
+      "loss": 1.3154,
+      "step": 9948
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4405407905578613,
+      "learning_rate": 4.737272006344172e-05,
+      "loss": 1.3323,
+      "step": 9949
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9799748659133911,
+      "learning_rate": 4.737245572297119e-05,
+      "loss": 1.6342,
+      "step": 9950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6266623735427856,
+      "learning_rate": 4.737219138250066e-05,
+      "loss": 1.3965,
+      "step": 9951
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.54586660861969,
+      "learning_rate": 4.737192704203014e-05,
+      "loss": 1.3424,
+      "step": 9952
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5899617671966553,
+      "learning_rate": 4.737166270155961e-05,
+      "loss": 2.2766,
+      "step": 9953
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5944420099258423,
+      "learning_rate": 4.737139836108909e-05,
+      "loss": 1.4141,
+      "step": 9954
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.667982816696167,
+      "learning_rate": 4.7371134020618555e-05,
+      "loss": 1.4109,
+      "step": 9955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8161317110061646,
+      "learning_rate": 4.7370869680148035e-05,
+      "loss": 1.5069,
+      "step": 9956
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8665367364883423,
+      "learning_rate": 4.737060533967751e-05,
+      "loss": 1.4489,
+      "step": 9957
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8821543455123901,
+      "learning_rate": 4.737034099920698e-05,
+      "loss": 1.2745,
+      "step": 9958
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8758574724197388,
+      "learning_rate": 4.7370076658736456e-05,
+      "loss": 1.3273,
+      "step": 9959
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8986586332321167,
+      "learning_rate": 4.736981231826593e-05,
+      "loss": 1.2725,
+      "step": 9960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5312796831130981,
+      "learning_rate": 4.73695479777954e-05,
+      "loss": 1.2856,
+      "step": 9961
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5312610864639282,
+      "learning_rate": 4.7369283637324877e-05,
+      "loss": 1.8998,
+      "step": 9962
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8763679265975952,
+      "learning_rate": 4.736901929685435e-05,
+      "loss": 1.4137,
+      "step": 9963
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6431235074996948,
+      "learning_rate": 4.7368754956383824e-05,
+      "loss": 1.3756,
+      "step": 9964
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.061779737472534,
+      "learning_rate": 4.73684906159133e-05,
+      "loss": 1.3683,
+      "step": 9965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5862528085708618,
+      "learning_rate": 4.736822627544277e-05,
+      "loss": 1.3191,
+      "step": 9966
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5680044889450073,
+      "learning_rate": 4.736796193497225e-05,
+      "loss": 1.2912,
+      "step": 9967
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1376218795776367,
+      "learning_rate": 4.736769759450172e-05,
+      "loss": 2.2511,
+      "step": 9968
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7221461534500122,
+      "learning_rate": 4.73674332540312e-05,
+      "loss": 1.3885,
+      "step": 9969
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7106940746307373,
+      "learning_rate": 4.7367168913560665e-05,
+      "loss": 1.4289,
+      "step": 9970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3241733312606812,
+      "learning_rate": 4.7366904573090145e-05,
+      "loss": 1.2242,
+      "step": 9971
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6954929828643799,
+      "learning_rate": 4.736664023261962e-05,
+      "loss": 1.3195,
+      "step": 9972
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6640211343765259,
+      "learning_rate": 4.736637589214909e-05,
+      "loss": 1.275,
+      "step": 9973
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9594188928604126,
+      "learning_rate": 4.7366111551678566e-05,
+      "loss": 1.4658,
+      "step": 9974
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0324013233184814,
+      "learning_rate": 4.736584721120804e-05,
+      "loss": 1.3227,
+      "step": 9975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6470248699188232,
+      "learning_rate": 4.736558287073751e-05,
+      "loss": 1.3699,
+      "step": 9976
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7017545700073242,
+      "learning_rate": 4.736531853026699e-05,
+      "loss": 1.4542,
+      "step": 9977
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.578250527381897,
+      "learning_rate": 4.736505418979646e-05,
+      "loss": 1.4054,
+      "step": 9978
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.596936821937561,
+      "learning_rate": 4.7364789849325934e-05,
+      "loss": 1.351,
+      "step": 9979
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4739781618118286,
+      "learning_rate": 4.736452550885541e-05,
+      "loss": 1.3366,
+      "step": 9980
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.702012062072754,
+      "learning_rate": 4.736426116838488e-05,
+      "loss": 1.3333,
+      "step": 9981
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6594555377960205,
+      "learning_rate": 4.736399682791436e-05,
+      "loss": 1.3062,
+      "step": 9982
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9955737590789795,
+      "learning_rate": 4.736373248744383e-05,
+      "loss": 1.3695,
+      "step": 9983
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2454512119293213,
+      "learning_rate": 4.736346814697331e-05,
+      "loss": 1.191,
+      "step": 9984
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7376463413238525,
+      "learning_rate": 4.7363203806502775e-05,
+      "loss": 1.4648,
+      "step": 9985
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2629568576812744,
+      "learning_rate": 4.736293946603225e-05,
+      "loss": 1.4372,
+      "step": 9986
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8617355823516846,
+      "learning_rate": 4.736267512556173e-05,
+      "loss": 2.1636,
+      "step": 9987
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8033852577209473,
+      "learning_rate": 4.7362410785091196e-05,
+      "loss": 1.3499,
+      "step": 9988
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5992451906204224,
+      "learning_rate": 4.7362146444620677e-05,
+      "loss": 1.4198,
+      "step": 9989
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4736911058425903,
+      "learning_rate": 4.736188210415014e-05,
+      "loss": 1.3712,
+      "step": 9990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5294337272644043,
+      "learning_rate": 4.7361617763679624e-05,
+      "loss": 1.4005,
+      "step": 9991
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.055217981338501,
+      "learning_rate": 4.73613534232091e-05,
+      "loss": 1.468,
+      "step": 9992
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.488048553466797,
+      "learning_rate": 4.736108908273857e-05,
+      "loss": 1.4573,
+      "step": 9993
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6704764366149902,
+      "learning_rate": 4.7360824742268044e-05,
+      "loss": 1.5525,
+      "step": 9994
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2699310779571533,
+      "learning_rate": 4.736056040179752e-05,
+      "loss": 1.4983,
+      "step": 9995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8948725461959839,
+      "learning_rate": 4.736029606132699e-05,
+      "loss": 1.2756,
+      "step": 9996
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3364415168762207,
+      "learning_rate": 4.7360031720856465e-05,
+      "loss": 1.342,
+      "step": 9997
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9439038038253784,
+      "learning_rate": 4.735976738038594e-05,
+      "loss": 1.4472,
+      "step": 9998
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5459201335906982,
+      "learning_rate": 4.735950303991541e-05,
+      "loss": 1.3453,
+      "step": 9999
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4851800203323364,
+      "learning_rate": 4.7359238699444886e-05,
+      "loss": 2.2663,
+      "step": 10000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7482422590255737,
+      "learning_rate": 4.735897435897436e-05,
+      "loss": 1.3773,
+      "step": 10001
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.742088794708252,
+      "learning_rate": 4.735871001850384e-05,
+      "loss": 1.5661,
+      "step": 10002
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0545647144317627,
+      "learning_rate": 4.7358445678033307e-05,
+      "loss": 1.4425,
+      "step": 10003
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6289111375808716,
+      "learning_rate": 4.735818133756279e-05,
+      "loss": 1.4398,
+      "step": 10004
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6181421279907227,
+      "learning_rate": 4.7357916997092254e-05,
+      "loss": 1.3286,
+      "step": 10005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.941129207611084,
+      "learning_rate": 4.7357652656621734e-05,
+      "loss": 1.4069,
+      "step": 10006
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8451460599899292,
+      "learning_rate": 4.735738831615121e-05,
+      "loss": 1.3965,
+      "step": 10007
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265411615371704,
+      "learning_rate": 4.735712397568068e-05,
+      "loss": 1.2872,
+      "step": 10008
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.97210431098938,
+      "learning_rate": 4.7356859635210155e-05,
+      "loss": 1.4002,
+      "step": 10009
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5393686294555664,
+      "learning_rate": 4.735659529473963e-05,
+      "loss": 1.2627,
+      "step": 10010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7660315036773682,
+      "learning_rate": 4.73563309542691e-05,
+      "loss": 1.4795,
+      "step": 10011
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.103408098220825,
+      "learning_rate": 4.7356066613798575e-05,
+      "loss": 1.5822,
+      "step": 10012
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7212828397750854,
+      "learning_rate": 4.735580227332805e-05,
+      "loss": 1.3919,
+      "step": 10013
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1228764057159424,
+      "learning_rate": 4.735553793285752e-05,
+      "loss": 1.421,
+      "step": 10014
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2526755332946777,
+      "learning_rate": 4.7355273592386996e-05,
+      "loss": 1.4984,
+      "step": 10015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7350298166275024,
+      "learning_rate": 4.735500925191647e-05,
+      "loss": 1.3266,
+      "step": 10016
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.508721351623535,
+      "learning_rate": 4.735474491144595e-05,
+      "loss": 1.4318,
+      "step": 10017
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1454341411590576,
+      "learning_rate": 4.735448057097542e-05,
+      "loss": 1.4447,
+      "step": 10018
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4322869777679443,
+      "learning_rate": 4.73542162305049e-05,
+      "loss": 1.4878,
+      "step": 10019
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7383692264556885,
+      "learning_rate": 4.7353951890034364e-05,
+      "loss": 1.3545,
+      "step": 10020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.030369281768799,
+      "learning_rate": 4.7353687549563844e-05,
+      "loss": 1.5195,
+      "step": 10021
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4040000438690186,
+      "learning_rate": 4.735342320909332e-05,
+      "loss": 1.4011,
+      "step": 10022
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9753350019454956,
+      "learning_rate": 4.7353158868622785e-05,
+      "loss": 1.3692,
+      "step": 10023
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5078972578048706,
+      "learning_rate": 4.7352894528152265e-05,
+      "loss": 1.2897,
+      "step": 10024
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.445841670036316,
+      "learning_rate": 4.735263018768173e-05,
+      "loss": 1.2215,
+      "step": 10025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.704224944114685,
+      "learning_rate": 4.735236584721121e-05,
+      "loss": 1.3285,
+      "step": 10026
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.085501194000244,
+      "learning_rate": 4.7352101506740686e-05,
+      "loss": 1.6584,
+      "step": 10027
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4573901891708374,
+      "learning_rate": 4.735183716627016e-05,
+      "loss": 1.321,
+      "step": 10028
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0376784801483154,
+      "learning_rate": 4.735157282579963e-05,
+      "loss": 1.5502,
+      "step": 10029
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.144972324371338,
+      "learning_rate": 4.7351308485329106e-05,
+      "loss": 1.4381,
+      "step": 10030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5779601335525513,
+      "learning_rate": 4.735104414485858e-05,
+      "loss": 1.3481,
+      "step": 10031
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9369192123413086,
+      "learning_rate": 4.7350779804388054e-05,
+      "loss": 1.3141,
+      "step": 10032
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5752860307693481,
+      "learning_rate": 4.735051546391753e-05,
+      "loss": 1.3711,
+      "step": 10033
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4384727478027344,
+      "learning_rate": 4.7350251123447e-05,
+      "loss": 2.034,
+      "step": 10034
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.024123430252075,
+      "learning_rate": 4.7349986782976474e-05,
+      "loss": 1.4057,
+      "step": 10035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8715168237686157,
+      "learning_rate": 4.734972244250595e-05,
+      "loss": 1.5304,
+      "step": 10036
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.215765953063965,
+      "learning_rate": 4.734945810203543e-05,
+      "loss": 1.5909,
+      "step": 10037
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8104513883590698,
+      "learning_rate": 4.7349193761564895e-05,
+      "loss": 1.3787,
+      "step": 10038
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.832065463066101,
+      "learning_rate": 4.7348929421094375e-05,
+      "loss": 1.4731,
+      "step": 10039
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.443320393562317,
+      "learning_rate": 4.734866508062384e-05,
+      "loss": 1.3555,
+      "step": 10040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7392791509628296,
+      "learning_rate": 4.734840074015332e-05,
+      "loss": 1.4722,
+      "step": 10041
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359008312225342,
+      "learning_rate": 4.7348136399682796e-05,
+      "loss": 1.4396,
+      "step": 10042
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9021072387695312,
+      "learning_rate": 4.734787205921227e-05,
+      "loss": 1.5071,
+      "step": 10043
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3011693954467773,
+      "learning_rate": 4.734760771874174e-05,
+      "loss": 1.1994,
+      "step": 10044
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6693934202194214,
+      "learning_rate": 4.734734337827122e-05,
+      "loss": 1.4159,
+      "step": 10045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9674475193023682,
+      "learning_rate": 4.734707903780069e-05,
+      "loss": 2.231,
+      "step": 10046
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2610862255096436,
+      "learning_rate": 4.7346814697330164e-05,
+      "loss": 1.2253,
+      "step": 10047
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.756037950515747,
+      "learning_rate": 4.734655035685964e-05,
+      "loss": 1.4463,
+      "step": 10048
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8814820051193237,
+      "learning_rate": 4.734628601638911e-05,
+      "loss": 1.255,
+      "step": 10049
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1036298274993896,
+      "learning_rate": 4.7346021675918585e-05,
+      "loss": 1.4649,
+      "step": 10050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1827449798583984,
+      "learning_rate": 4.734575733544806e-05,
+      "loss": 1.4735,
+      "step": 10051
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.94602370262146,
+      "learning_rate": 4.734549299497754e-05,
+      "loss": 1.5305,
+      "step": 10052
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8408010005950928,
+      "learning_rate": 4.7345228654507005e-05,
+      "loss": 1.3538,
+      "step": 10053
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6653025150299072,
+      "learning_rate": 4.7344964314036486e-05,
+      "loss": 1.5445,
+      "step": 10054
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9107884168624878,
+      "learning_rate": 4.734469997356595e-05,
+      "loss": 1.2358,
+      "step": 10055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7745039463043213,
+      "learning_rate": 4.734443563309543e-05,
+      "loss": 1.3208,
+      "step": 10056
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.167550563812256,
+      "learning_rate": 4.7344171292624906e-05,
+      "loss": 1.4122,
+      "step": 10057
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.216935873031616,
+      "learning_rate": 4.734390695215437e-05,
+      "loss": 1.4343,
+      "step": 10058
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9534286260604858,
+      "learning_rate": 4.7343642611683854e-05,
+      "loss": 1.5275,
+      "step": 10059
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3277668952941895,
+      "learning_rate": 4.734337827121332e-05,
+      "loss": 1.2917,
+      "step": 10060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.827791452407837,
+      "learning_rate": 4.73431139307428e-05,
+      "loss": 1.3502,
+      "step": 10061
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6411951780319214,
+      "learning_rate": 4.734284959027227e-05,
+      "loss": 1.426,
+      "step": 10062
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7451083660125732,
+      "learning_rate": 4.734258524980175e-05,
+      "loss": 1.5163,
+      "step": 10063
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.114011764526367,
+      "learning_rate": 4.734232090933122e-05,
+      "loss": 1.5293,
+      "step": 10064
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1005797386169434,
+      "learning_rate": 4.7342056568860695e-05,
+      "loss": 1.7055,
+      "step": 10065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6073744297027588,
+      "learning_rate": 4.734179222839017e-05,
+      "loss": 1.3554,
+      "step": 10066
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7459098100662231,
+      "learning_rate": 4.734152788791964e-05,
+      "loss": 1.3077,
+      "step": 10067
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6445344686508179,
+      "learning_rate": 4.7341263547449116e-05,
+      "loss": 1.4246,
+      "step": 10068
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9051603078842163,
+      "learning_rate": 4.734099920697859e-05,
+      "loss": 1.4696,
+      "step": 10069
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9383474588394165,
+      "learning_rate": 4.734073486650806e-05,
+      "loss": 1.4639,
+      "step": 10070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9625314474105835,
+      "learning_rate": 4.7340470526037536e-05,
+      "loss": 1.4681,
+      "step": 10071
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.669379472732544,
+      "learning_rate": 4.734020618556701e-05,
+      "loss": 1.3753,
+      "step": 10072
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1911818981170654,
+      "learning_rate": 4.7339941845096484e-05,
+      "loss": 2.6189,
+      "step": 10073
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9213889837265015,
+      "learning_rate": 4.7339677504625964e-05,
+      "loss": 1.459,
+      "step": 10074
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9291951656341553,
+      "learning_rate": 4.733941316415543e-05,
+      "loss": 1.3367,
+      "step": 10075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7249926328659058,
+      "learning_rate": 4.733914882368491e-05,
+      "loss": 1.3479,
+      "step": 10076
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8180283308029175,
+      "learning_rate": 4.733888448321438e-05,
+      "loss": 1.4237,
+      "step": 10077
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.164137363433838,
+      "learning_rate": 4.733862014274386e-05,
+      "loss": 1.559,
+      "step": 10078
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9499688148498535,
+      "learning_rate": 4.733835580227333e-05,
+      "loss": 1.3201,
+      "step": 10079
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.735649585723877,
+      "learning_rate": 4.7338091461802805e-05,
+      "loss": 1.3814,
+      "step": 10080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4086787700653076,
+      "learning_rate": 4.733782712133228e-05,
+      "loss": 1.6705,
+      "step": 10081
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9183377027511597,
+      "learning_rate": 4.733756278086175e-05,
+      "loss": 1.538,
+      "step": 10082
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9135535955429077,
+      "learning_rate": 4.7337298440391226e-05,
+      "loss": 1.2715,
+      "step": 10083
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.510125160217285,
+      "learning_rate": 4.73370340999207e-05,
+      "loss": 1.4124,
+      "step": 10084
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4230250120162964,
+      "learning_rate": 4.733676975945017e-05,
+      "loss": 1.3735,
+      "step": 10085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8838422298431396,
+      "learning_rate": 4.733650541897965e-05,
+      "loss": 1.4378,
+      "step": 10086
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6464085578918457,
+      "learning_rate": 4.733624107850912e-05,
+      "loss": 1.4214,
+      "step": 10087
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.123534917831421,
+      "learning_rate": 4.7335976738038594e-05,
+      "loss": 1.3034,
+      "step": 10088
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5761252641677856,
+      "learning_rate": 4.7335712397568074e-05,
+      "loss": 1.4448,
+      "step": 10089
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7235727310180664,
+      "learning_rate": 4.733544805709754e-05,
+      "loss": 1.3478,
+      "step": 10090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0182838439941406,
+      "learning_rate": 4.733518371662702e-05,
+      "loss": 1.3456,
+      "step": 10091
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6354429721832275,
+      "learning_rate": 4.733491937615649e-05,
+      "loss": 1.3513,
+      "step": 10092
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6656776666641235,
+      "learning_rate": 4.733465503568597e-05,
+      "loss": 1.4068,
+      "step": 10093
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0773801803588867,
+      "learning_rate": 4.733439069521544e-05,
+      "loss": 1.2444,
+      "step": 10094
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6267186403274536,
+      "learning_rate": 4.733412635474491e-05,
+      "loss": 1.2716,
+      "step": 10095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9381132125854492,
+      "learning_rate": 4.733386201427439e-05,
+      "loss": 1.6234,
+      "step": 10096
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.216766119003296,
+      "learning_rate": 4.7333597673803856e-05,
+      "loss": 1.5694,
+      "step": 10097
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4039671421051025,
+      "learning_rate": 4.7333333333333336e-05,
+      "loss": 1.4349,
+      "step": 10098
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9766474962234497,
+      "learning_rate": 4.733306899286281e-05,
+      "loss": 1.3689,
+      "step": 10099
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4925175905227661,
+      "learning_rate": 4.7332804652392284e-05,
+      "loss": 1.277,
+      "step": 10100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2892194986343384,
+      "learning_rate": 4.733254031192176e-05,
+      "loss": 1.3797,
+      "step": 10101
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4313435554504395,
+      "learning_rate": 4.733227597145123e-05,
+      "loss": 1.5342,
+      "step": 10102
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6041057109832764,
+      "learning_rate": 4.7332011630980704e-05,
+      "loss": 1.2846,
+      "step": 10103
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1658101081848145,
+      "learning_rate": 4.733174729051018e-05,
+      "loss": 1.4494,
+      "step": 10104
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6563830375671387,
+      "learning_rate": 4.733148295003965e-05,
+      "loss": 1.4744,
+      "step": 10105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5434707403182983,
+      "learning_rate": 4.7331218609569125e-05,
+      "loss": 1.3397,
+      "step": 10106
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6971853971481323,
+      "learning_rate": 4.73309542690986e-05,
+      "loss": 1.6352,
+      "step": 10107
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5117785930633545,
+      "learning_rate": 4.733068992862807e-05,
+      "loss": 1.3459,
+      "step": 10108
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6689871549606323,
+      "learning_rate": 4.733042558815755e-05,
+      "loss": 1.5192,
+      "step": 10109
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1552951335906982,
+      "learning_rate": 4.733016124768702e-05,
+      "loss": 1.5091,
+      "step": 10110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5377274751663208,
+      "learning_rate": 4.73298969072165e-05,
+      "loss": 1.185,
+      "step": 10111
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7208994626998901,
+      "learning_rate": 4.7329632566745966e-05,
+      "loss": 1.2473,
+      "step": 10112
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.600325584411621,
+      "learning_rate": 4.732936822627545e-05,
+      "loss": 1.5047,
+      "step": 10113
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.817878246307373,
+      "learning_rate": 4.732910388580492e-05,
+      "loss": 1.3205,
+      "step": 10114
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7408910989761353,
+      "learning_rate": 4.7328839545334394e-05,
+      "loss": 1.5081,
+      "step": 10115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9034194946289062,
+      "learning_rate": 4.732857520486387e-05,
+      "loss": 1.3671,
+      "step": 10116
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9011071920394897,
+      "learning_rate": 4.732831086439334e-05,
+      "loss": 1.2984,
+      "step": 10117
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1154496669769287,
+      "learning_rate": 4.7328046523922815e-05,
+      "loss": 1.4334,
+      "step": 10118
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6750315427780151,
+      "learning_rate": 4.732778218345229e-05,
+      "loss": 1.2211,
+      "step": 10119
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7522666454315186,
+      "learning_rate": 4.732751784298176e-05,
+      "loss": 1.4134,
+      "step": 10120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5587605237960815,
+      "learning_rate": 4.7327253502511235e-05,
+      "loss": 1.1114,
+      "step": 10121
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0515871047973633,
+      "learning_rate": 4.732698916204071e-05,
+      "loss": 1.5686,
+      "step": 10122
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6645697355270386,
+      "learning_rate": 4.732672482157018e-05,
+      "loss": 1.3236,
+      "step": 10123
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5599424839019775,
+      "learning_rate": 4.732646048109966e-05,
+      "loss": 1.6622,
+      "step": 10124
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8635262250900269,
+      "learning_rate": 4.732619614062913e-05,
+      "loss": 1.5562,
+      "step": 10125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6234220266342163,
+      "learning_rate": 4.732593180015861e-05,
+      "loss": 1.3301,
+      "step": 10126
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8949756622314453,
+      "learning_rate": 4.732566745968808e-05,
+      "loss": 1.3759,
+      "step": 10127
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5053027868270874,
+      "learning_rate": 4.732540311921756e-05,
+      "loss": 1.317,
+      "step": 10128
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5518815517425537,
+      "learning_rate": 4.732513877874703e-05,
+      "loss": 1.3822,
+      "step": 10129
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3915935754776,
+      "learning_rate": 4.73248744382765e-05,
+      "loss": 1.2864,
+      "step": 10130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9647470712661743,
+      "learning_rate": 4.732461009780598e-05,
+      "loss": 1.6301,
+      "step": 10131
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7163336277008057,
+      "learning_rate": 4.7324345757335445e-05,
+      "loss": 1.2859,
+      "step": 10132
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.998532772064209,
+      "learning_rate": 4.7324081416864925e-05,
+      "loss": 1.4062,
+      "step": 10133
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2027764320373535,
+      "learning_rate": 4.73238170763944e-05,
+      "loss": 1.5286,
+      "step": 10134
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8810639381408691,
+      "learning_rate": 4.732355273592387e-05,
+      "loss": 1.4925,
+      "step": 10135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5879875421524048,
+      "learning_rate": 4.7323288395453346e-05,
+      "loss": 1.3074,
+      "step": 10136
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7180819511413574,
+      "learning_rate": 4.732302405498282e-05,
+      "loss": 1.3816,
+      "step": 10137
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5940160751342773,
+      "learning_rate": 4.732275971451229e-05,
+      "loss": 1.4129,
+      "step": 10138
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0372164249420166,
+      "learning_rate": 4.732249537404177e-05,
+      "loss": 1.2834,
+      "step": 10139
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8187106847763062,
+      "learning_rate": 4.732223103357124e-05,
+      "loss": 1.5718,
+      "step": 10140
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.508208751678467,
+      "learning_rate": 4.7321966693100714e-05,
+      "loss": 1.5133,
+      "step": 10141
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7059389352798462,
+      "learning_rate": 4.732170235263019e-05,
+      "loss": 1.4792,
+      "step": 10142
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9003522396087646,
+      "learning_rate": 4.732143801215966e-05,
+      "loss": 1.3075,
+      "step": 10143
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8249545097351074,
+      "learning_rate": 4.732117367168914e-05,
+      "loss": 1.3972,
+      "step": 10144
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7762795686721802,
+      "learning_rate": 4.732090933121861e-05,
+      "loss": 1.5454,
+      "step": 10145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0220890045166016,
+      "learning_rate": 4.732064499074809e-05,
+      "loss": 1.2454,
+      "step": 10146
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6809145212173462,
+      "learning_rate": 4.7320380650277555e-05,
+      "loss": 1.4124,
+      "step": 10147
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.137601852416992,
+      "learning_rate": 4.7320116309807035e-05,
+      "loss": 1.5716,
+      "step": 10148
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3255852460861206,
+      "learning_rate": 4.731985196933651e-05,
+      "loss": 1.3268,
+      "step": 10149
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8525292873382568,
+      "learning_rate": 4.731958762886598e-05,
+      "loss": 1.3348,
+      "step": 10150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.06339430809021,
+      "learning_rate": 4.7319323288395456e-05,
+      "loss": 1.4806,
+      "step": 10151
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7767349481582642,
+      "learning_rate": 4.731905894792493e-05,
+      "loss": 1.2523,
+      "step": 10152
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.322432041168213,
+      "learning_rate": 4.73187946074544e-05,
+      "loss": 1.2511,
+      "step": 10153
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7260668277740479,
+      "learning_rate": 4.731853026698388e-05,
+      "loss": 2.2449,
+      "step": 10154
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6118112802505493,
+      "learning_rate": 4.731826592651335e-05,
+      "loss": 1.3886,
+      "step": 10155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7572287321090698,
+      "learning_rate": 4.7318001586042824e-05,
+      "loss": 1.3784,
+      "step": 10156
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.604522705078125,
+      "learning_rate": 4.73177372455723e-05,
+      "loss": 1.3001,
+      "step": 10157
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8334053754806519,
+      "learning_rate": 4.731747290510177e-05,
+      "loss": 1.3928,
+      "step": 10158
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6371879577636719,
+      "learning_rate": 4.731720856463125e-05,
+      "loss": 1.2859,
+      "step": 10159
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5914610624313354,
+      "learning_rate": 4.731694422416072e-05,
+      "loss": 1.3183,
+      "step": 10160
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5751904249191284,
+      "learning_rate": 4.73166798836902e-05,
+      "loss": 1.3756,
+      "step": 10161
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8761283159255981,
+      "learning_rate": 4.7316415543219665e-05,
+      "loss": 1.5338,
+      "step": 10162
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5608093738555908,
+      "learning_rate": 4.7316151202749146e-05,
+      "loss": 1.3811,
+      "step": 10163
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5758131742477417,
+      "learning_rate": 4.731588686227862e-05,
+      "loss": 1.4161,
+      "step": 10164
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.666399359703064,
+      "learning_rate": 4.731562252180809e-05,
+      "loss": 1.3245,
+      "step": 10165
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.411684036254883,
+      "learning_rate": 4.7315358181337566e-05,
+      "loss": 1.4608,
+      "step": 10166
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.525875449180603,
+      "learning_rate": 4.731509384086703e-05,
+      "loss": 1.3258,
+      "step": 10167
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.333989143371582,
+      "learning_rate": 4.7314829500396513e-05,
+      "loss": 1.0071,
+      "step": 10168
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.348954439163208,
+      "learning_rate": 4.731456515992599e-05,
+      "loss": 1.6085,
+      "step": 10169
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.752007246017456,
+      "learning_rate": 4.731430081945546e-05,
+      "loss": 1.5034,
+      "step": 10170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1418795585632324,
+      "learning_rate": 4.7314036478984934e-05,
+      "loss": 1.4619,
+      "step": 10171
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.422283411026001,
+      "learning_rate": 4.731377213851441e-05,
+      "loss": 1.3279,
+      "step": 10172
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7205525636672974,
+      "learning_rate": 4.731350779804388e-05,
+      "loss": 1.542,
+      "step": 10173
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.146664619445801,
+      "learning_rate": 4.731324345757336e-05,
+      "loss": 1.2724,
+      "step": 10174
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9512701034545898,
+      "learning_rate": 4.731297911710283e-05,
+      "loss": 1.4271,
+      "step": 10175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.829260230064392,
+      "learning_rate": 4.73127147766323e-05,
+      "loss": 1.4808,
+      "step": 10176
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.165235996246338,
+      "learning_rate": 4.7312450436161776e-05,
+      "loss": 1.4084,
+      "step": 10177
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6922507286071777,
+      "learning_rate": 4.731218609569125e-05,
+      "loss": 1.241,
+      "step": 10178
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5182067155838013,
+      "learning_rate": 4.731192175522073e-05,
+      "loss": 1.2709,
+      "step": 10179
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2321529388427734,
+      "learning_rate": 4.7311657414750196e-05,
+      "loss": 1.5667,
+      "step": 10180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7259321212768555,
+      "learning_rate": 4.731139307427968e-05,
+      "loss": 1.3022,
+      "step": 10181
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9143580198287964,
+      "learning_rate": 4.7311128733809143e-05,
+      "loss": 1.3405,
+      "step": 10182
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9728938341140747,
+      "learning_rate": 4.7310864393338624e-05,
+      "loss": 1.503,
+      "step": 10183
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.381803274154663,
+      "learning_rate": 4.73106000528681e-05,
+      "loss": 1.2799,
+      "step": 10184
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.026348114013672,
+      "learning_rate": 4.731033571239757e-05,
+      "loss": 1.3616,
+      "step": 10185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.047316551208496,
+      "learning_rate": 4.7310071371927045e-05,
+      "loss": 1.4426,
+      "step": 10186
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0019426345825195,
+      "learning_rate": 4.730980703145652e-05,
+      "loss": 1.5649,
+      "step": 10187
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4779540300369263,
+      "learning_rate": 4.730954269098599e-05,
+      "loss": 1.3696,
+      "step": 10188
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6856731176376343,
+      "learning_rate": 4.7309278350515465e-05,
+      "loss": 1.4201,
+      "step": 10189
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2626380920410156,
+      "learning_rate": 4.730901401004494e-05,
+      "loss": 1.3836,
+      "step": 10190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4937381744384766,
+      "learning_rate": 4.730874966957441e-05,
+      "loss": 1.3569,
+      "step": 10191
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0289878845214844,
+      "learning_rate": 4.7308485329103886e-05,
+      "loss": 1.6526,
+      "step": 10192
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3429036140441895,
+      "learning_rate": 4.730822098863336e-05,
+      "loss": 1.5184,
+      "step": 10193
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7964907884597778,
+      "learning_rate": 4.730795664816284e-05,
+      "loss": 1.3437,
+      "step": 10194
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.215801477432251,
+      "learning_rate": 4.730769230769231e-05,
+      "loss": 1.5421,
+      "step": 10195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5129457712173462,
+      "learning_rate": 4.730742796722179e-05,
+      "loss": 1.3371,
+      "step": 10196
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5767241716384888,
+      "learning_rate": 4.7307163626751254e-05,
+      "loss": 1.3874,
+      "step": 10197
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.823358416557312,
+      "learning_rate": 4.7306899286280734e-05,
+      "loss": 1.3648,
+      "step": 10198
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6688623428344727,
+      "learning_rate": 4.730663494581021e-05,
+      "loss": 1.4736,
+      "step": 10199
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5182656049728394,
+      "learning_rate": 4.730637060533968e-05,
+      "loss": 1.3255,
+      "step": 10200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3722749948501587,
+      "learning_rate": 4.7306106264869155e-05,
+      "loss": 1.3011,
+      "step": 10201
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.105813503265381,
+      "learning_rate": 4.730584192439862e-05,
+      "loss": 1.6112,
+      "step": 10202
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9995753765106201,
+      "learning_rate": 4.73055775839281e-05,
+      "loss": 1.3335,
+      "step": 10203
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1510024070739746,
+      "learning_rate": 4.7305313243457576e-05,
+      "loss": 1.4193,
+      "step": 10204
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9980623722076416,
+      "learning_rate": 4.730504890298705e-05,
+      "loss": 1.6002,
+      "step": 10205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.282052516937256,
+      "learning_rate": 4.730478456251652e-05,
+      "loss": 1.6858,
+      "step": 10206
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.206758975982666,
+      "learning_rate": 4.7304520222045996e-05,
+      "loss": 1.413,
+      "step": 10207
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4326927661895752,
+      "learning_rate": 4.730425588157547e-05,
+      "loss": 1.3192,
+      "step": 10208
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8173022270202637,
+      "learning_rate": 4.730399154110495e-05,
+      "loss": 1.3774,
+      "step": 10209
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.868391990661621,
+      "learning_rate": 4.730372720063442e-05,
+      "loss": 1.4718,
+      "step": 10210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4083645343780518,
+      "learning_rate": 4.73034628601639e-05,
+      "loss": 1.3081,
+      "step": 10211
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8605308532714844,
+      "learning_rate": 4.7303198519693364e-05,
+      "loss": 1.4843,
+      "step": 10212
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.572526216506958,
+      "learning_rate": 4.730293417922284e-05,
+      "loss": 1.5238,
+      "step": 10213
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.784085750579834,
+      "learning_rate": 4.730266983875232e-05,
+      "loss": 2.0118,
+      "step": 10214
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5517346858978271,
+      "learning_rate": 4.7302405498281785e-05,
+      "loss": 1.4164,
+      "step": 10215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4487239122390747,
+      "learning_rate": 4.7302141157811265e-05,
+      "loss": 1.3418,
+      "step": 10216
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.693389654159546,
+      "learning_rate": 4.730187681734073e-05,
+      "loss": 1.389,
+      "step": 10217
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7036404609680176,
+      "learning_rate": 4.730161247687021e-05,
+      "loss": 1.2758,
+      "step": 10218
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0599095821380615,
+      "learning_rate": 4.7301348136399686e-05,
+      "loss": 1.4566,
+      "step": 10219
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.704804539680481,
+      "learning_rate": 4.730108379592916e-05,
+      "loss": 1.3169,
+      "step": 10220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8471828699111938,
+      "learning_rate": 4.730081945545863e-05,
+      "loss": 1.796,
+      "step": 10221
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6424106359481812,
+      "learning_rate": 4.730055511498811e-05,
+      "loss": 1.4314,
+      "step": 10222
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8558167219161987,
+      "learning_rate": 4.730029077451758e-05,
+      "loss": 1.406,
+      "step": 10223
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9279171228408813,
+      "learning_rate": 4.7300026434047054e-05,
+      "loss": 1.5148,
+      "step": 10224
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.167745590209961,
+      "learning_rate": 4.729976209357653e-05,
+      "loss": 2.1011,
+      "step": 10225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7299306392669678,
+      "learning_rate": 4.7299497753106e-05,
+      "loss": 1.3328,
+      "step": 10226
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5155043601989746,
+      "learning_rate": 4.7299233412635474e-05,
+      "loss": 1.2941,
+      "step": 10227
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7629623413085938,
+      "learning_rate": 4.729896907216495e-05,
+      "loss": 1.4192,
+      "step": 10228
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.985164999961853,
+      "learning_rate": 4.729870473169443e-05,
+      "loss": 1.4062,
+      "step": 10229
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6351747512817383,
+      "learning_rate": 4.7298440391223895e-05,
+      "loss": 1.3773,
+      "step": 10230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3404701948165894,
+      "learning_rate": 4.7298176050753376e-05,
+      "loss": 1.271,
+      "step": 10231
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.635457754135132,
+      "learning_rate": 4.729791171028284e-05,
+      "loss": 1.9647,
+      "step": 10232
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6749329566955566,
+      "learning_rate": 4.729764736981232e-05,
+      "loss": 1.4619,
+      "step": 10233
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6256440877914429,
+      "learning_rate": 4.7297383029341796e-05,
+      "loss": 1.4825,
+      "step": 10234
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4740973711013794,
+      "learning_rate": 4.729711868887127e-05,
+      "loss": 1.3757,
+      "step": 10235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.990746259689331,
+      "learning_rate": 4.7296854348400743e-05,
+      "loss": 1.5947,
+      "step": 10236
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3293890953063965,
+      "learning_rate": 4.729659000793022e-05,
+      "loss": 1.4572,
+      "step": 10237
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0088043212890625,
+      "learning_rate": 4.729632566745969e-05,
+      "loss": 1.5978,
+      "step": 10238
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8977930545806885,
+      "learning_rate": 4.7296061326989164e-05,
+      "loss": 1.3373,
+      "step": 10239
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7497485876083374,
+      "learning_rate": 4.729579698651864e-05,
+      "loss": 1.5221,
+      "step": 10240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.652174949645996,
+      "learning_rate": 4.729553264604811e-05,
+      "loss": 1.4897,
+      "step": 10241
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8466981649398804,
+      "learning_rate": 4.7295268305577585e-05,
+      "loss": 1.2969,
+      "step": 10242
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9964702129364014,
+      "learning_rate": 4.729500396510706e-05,
+      "loss": 1.4327,
+      "step": 10243
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2109103202819824,
+      "learning_rate": 4.729473962463654e-05,
+      "loss": 1.48,
+      "step": 10244
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6650429964065552,
+      "learning_rate": 4.7294475284166006e-05,
+      "loss": 1.4626,
+      "step": 10245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.637643337249756,
+      "learning_rate": 4.7294210943695486e-05,
+      "loss": 1.5392,
+      "step": 10246
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4974907636642456,
+      "learning_rate": 4.729394660322495e-05,
+      "loss": 1.44,
+      "step": 10247
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6650023460388184,
+      "learning_rate": 4.7293682262754426e-05,
+      "loss": 1.4579,
+      "step": 10248
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4701327085494995,
+      "learning_rate": 4.7293417922283907e-05,
+      "loss": 1.2776,
+      "step": 10249
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4383955001831055,
+      "learning_rate": 4.7293153581813373e-05,
+      "loss": 1.2739,
+      "step": 10250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.827238917350769,
+      "learning_rate": 4.7292889241342854e-05,
+      "loss": 1.1722,
+      "step": 10251
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7383170127868652,
+      "learning_rate": 4.729262490087232e-05,
+      "loss": 1.2397,
+      "step": 10252
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3610291481018066,
+      "learning_rate": 4.72923605604018e-05,
+      "loss": 1.2855,
+      "step": 10253
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9114571809768677,
+      "learning_rate": 4.7292096219931274e-05,
+      "loss": 1.2706,
+      "step": 10254
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0246241092681885,
+      "learning_rate": 4.729183187946075e-05,
+      "loss": 1.4402,
+      "step": 10255
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6715584993362427,
+      "learning_rate": 4.729156753899022e-05,
+      "loss": 1.3438,
+      "step": 10256
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5098960399627686,
+      "learning_rate": 4.7291303198519695e-05,
+      "loss": 1.3433,
+      "step": 10257
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5101481676101685,
+      "learning_rate": 4.729103885804917e-05,
+      "loss": 1.1807,
+      "step": 10258
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6786773204803467,
+      "learning_rate": 4.729077451757864e-05,
+      "loss": 1.2646,
+      "step": 10259
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9876186847686768,
+      "learning_rate": 4.7290510177108116e-05,
+      "loss": 1.4727,
+      "step": 10260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8469891548156738,
+      "learning_rate": 4.729024583663759e-05,
+      "loss": 1.4291,
+      "step": 10261
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8727916479110718,
+      "learning_rate": 4.728998149616706e-05,
+      "loss": 1.4698,
+      "step": 10262
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4931700229644775,
+      "learning_rate": 4.7289717155696537e-05,
+      "loss": 1.2519,
+      "step": 10263
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2359707355499268,
+      "learning_rate": 4.728945281522602e-05,
+      "loss": 2.0206,
+      "step": 10264
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4768357276916504,
+      "learning_rate": 4.7289188474755484e-05,
+      "loss": 1.4773,
+      "step": 10265
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.68362295627594,
+      "learning_rate": 4.7288924134284964e-05,
+      "loss": 1.3496,
+      "step": 10266
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5296727418899536,
+      "learning_rate": 4.728865979381443e-05,
+      "loss": 1.371,
+      "step": 10267
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.018198251724243,
+      "learning_rate": 4.728839545334391e-05,
+      "loss": 1.5286,
+      "step": 10268
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.86346435546875,
+      "learning_rate": 4.7288131112873385e-05,
+      "loss": 1.3,
+      "step": 10269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.289759397506714,
+      "learning_rate": 4.728786677240286e-05,
+      "loss": 1.3406,
+      "step": 10270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.379519462585449,
+      "learning_rate": 4.728760243193233e-05,
+      "loss": 1.3303,
+      "step": 10271
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3717305660247803,
+      "learning_rate": 4.7287338091461806e-05,
+      "loss": 1.4854,
+      "step": 10272
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.958165168762207,
+      "learning_rate": 4.728707375099128e-05,
+      "loss": 1.5161,
+      "step": 10273
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6233009099960327,
+      "learning_rate": 4.728680941052075e-05,
+      "loss": 1.2343,
+      "step": 10274
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1017279624938965,
+      "learning_rate": 4.7286545070050226e-05,
+      "loss": 1.5213,
+      "step": 10275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1141483783721924,
+      "learning_rate": 4.72862807295797e-05,
+      "loss": 1.3632,
+      "step": 10276
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6538915634155273,
+      "learning_rate": 4.728601638910917e-05,
+      "loss": 1.5087,
+      "step": 10277
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.562220811843872,
+      "learning_rate": 4.728575204863865e-05,
+      "loss": 1.3384,
+      "step": 10278
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.013667345046997,
+      "learning_rate": 4.728548770816813e-05,
+      "loss": 1.3429,
+      "step": 10279
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6761846542358398,
+      "learning_rate": 4.7285223367697594e-05,
+      "loss": 1.359,
+      "step": 10280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6559561491012573,
+      "learning_rate": 4.7284959027227074e-05,
+      "loss": 1.2658,
+      "step": 10281
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.195028781890869,
+      "learning_rate": 4.728469468675654e-05,
+      "loss": 1.3924,
+      "step": 10282
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9387296438217163,
+      "learning_rate": 4.728443034628602e-05,
+      "loss": 1.456,
+      "step": 10283
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4882738590240479,
+      "learning_rate": 4.7284166005815495e-05,
+      "loss": 1.6642,
+      "step": 10284
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7747563123703003,
+      "learning_rate": 4.728390166534496e-05,
+      "loss": 1.474,
+      "step": 10285
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.716718673706055,
+      "learning_rate": 4.728363732487444e-05,
+      "loss": 1.7087,
+      "step": 10286
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0898351669311523,
+      "learning_rate": 4.728337298440391e-05,
+      "loss": 1.5215,
+      "step": 10287
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6629678010940552,
+      "learning_rate": 4.728310864393339e-05,
+      "loss": 1.4072,
+      "step": 10288
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9460391998291016,
+      "learning_rate": 4.728284430346286e-05,
+      "loss": 1.5395,
+      "step": 10289
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.169358015060425,
+      "learning_rate": 4.7282579962992337e-05,
+      "loss": 1.3738,
+      "step": 10290
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7139114141464233,
+      "learning_rate": 4.728231562252181e-05,
+      "loss": 1.3049,
+      "step": 10291
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8901430368423462,
+      "learning_rate": 4.7282051282051284e-05,
+      "loss": 1.4713,
+      "step": 10292
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4825069904327393,
+      "learning_rate": 4.728178694158076e-05,
+      "loss": 1.2687,
+      "step": 10293
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3313370943069458,
+      "learning_rate": 4.728152260111023e-05,
+      "loss": 1.1784,
+      "step": 10294
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8918447494506836,
+      "learning_rate": 4.7281258260639704e-05,
+      "loss": 1.4161,
+      "step": 10295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5834171772003174,
+      "learning_rate": 4.728099392016918e-05,
+      "loss": 1.3957,
+      "step": 10296
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6150331497192383,
+      "learning_rate": 4.728072957969865e-05,
+      "loss": 1.4322,
+      "step": 10297
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5896739959716797,
+      "learning_rate": 4.7280465239228125e-05,
+      "loss": 1.204,
+      "step": 10298
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5352553129196167,
+      "learning_rate": 4.7280200898757605e-05,
+      "loss": 1.382,
+      "step": 10299
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6998783349990845,
+      "learning_rate": 4.727993655828707e-05,
+      "loss": 1.3483,
+      "step": 10300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5993611812591553,
+      "learning_rate": 4.727967221781655e-05,
+      "loss": 1.4858,
+      "step": 10301
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.394279956817627,
+      "learning_rate": 4.727940787734602e-05,
+      "loss": 1.2201,
+      "step": 10302
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7568260431289673,
+      "learning_rate": 4.72791435368755e-05,
+      "loss": 1.2633,
+      "step": 10303
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7995033264160156,
+      "learning_rate": 4.727887919640497e-05,
+      "loss": 1.5715,
+      "step": 10304
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0060079097747803,
+      "learning_rate": 4.727861485593445e-05,
+      "loss": 2.3874,
+      "step": 10305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.00986647605896,
+      "learning_rate": 4.727835051546392e-05,
+      "loss": 1.4279,
+      "step": 10306
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8798742294311523,
+      "learning_rate": 4.7278086174993394e-05,
+      "loss": 1.2774,
+      "step": 10307
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.495479941368103,
+      "learning_rate": 4.727782183452287e-05,
+      "loss": 1.3035,
+      "step": 10308
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.691676139831543,
+      "learning_rate": 4.727755749405234e-05,
+      "loss": 1.4799,
+      "step": 10309
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7405799627304077,
+      "learning_rate": 4.7277293153581815e-05,
+      "loss": 1.3998,
+      "step": 10310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.012793779373169,
+      "learning_rate": 4.727702881311129e-05,
+      "loss": 1.452,
+      "step": 10311
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2837592363357544,
+      "learning_rate": 4.727676447264076e-05,
+      "loss": 1.871,
+      "step": 10312
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3513107299804688,
+      "learning_rate": 4.7276500132170235e-05,
+      "loss": 1.3204,
+      "step": 10313
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5868947505950928,
+      "learning_rate": 4.7276235791699716e-05,
+      "loss": 1.36,
+      "step": 10314
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4458906650543213,
+      "learning_rate": 4.727597145122918e-05,
+      "loss": 1.393,
+      "step": 10315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6286065578460693,
+      "learning_rate": 4.727570711075866e-05,
+      "loss": 1.3982,
+      "step": 10316
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.769912838935852,
+      "learning_rate": 4.727544277028813e-05,
+      "loss": 1.4303,
+      "step": 10317
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.289189338684082,
+      "learning_rate": 4.727517842981761e-05,
+      "loss": 1.4865,
+      "step": 10318
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9931427240371704,
+      "learning_rate": 4.7274914089347084e-05,
+      "loss": 1.5396,
+      "step": 10319
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.696670651435852,
+      "learning_rate": 4.727464974887655e-05,
+      "loss": 1.414,
+      "step": 10320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.178126811981201,
+      "learning_rate": 4.727438540840603e-05,
+      "loss": 1.4707,
+      "step": 10321
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7953193187713623,
+      "learning_rate": 4.72741210679355e-05,
+      "loss": 1.3104,
+      "step": 10322
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5634137392044067,
+      "learning_rate": 4.727385672746498e-05,
+      "loss": 1.3118,
+      "step": 10323
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4525346755981445,
+      "learning_rate": 4.727359238699445e-05,
+      "loss": 1.3488,
+      "step": 10324
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.091660261154175,
+      "learning_rate": 4.7273328046523925e-05,
+      "loss": 1.415,
+      "step": 10325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6368964910507202,
+      "learning_rate": 4.72730637060534e-05,
+      "loss": 1.3412,
+      "step": 10326
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6152647733688354,
+      "learning_rate": 4.727279936558287e-05,
+      "loss": 1.4881,
+      "step": 10327
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8833041191101074,
+      "learning_rate": 4.7272535025112346e-05,
+      "loss": 1.75,
+      "step": 10328
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9365334510803223,
+      "learning_rate": 4.7272270684641826e-05,
+      "loss": 1.4987,
+      "step": 10329
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421715021133423,
+      "learning_rate": 4.727200634417129e-05,
+      "loss": 1.6669,
+      "step": 10330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6048656702041626,
+      "learning_rate": 4.7271742003700767e-05,
+      "loss": 1.3271,
+      "step": 10331
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6721795797348022,
+      "learning_rate": 4.727147766323024e-05,
+      "loss": 1.2331,
+      "step": 10332
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.647448778152466,
+      "learning_rate": 4.7271213322759714e-05,
+      "loss": 1.4045,
+      "step": 10333
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6160802841186523,
+      "learning_rate": 4.7270948982289194e-05,
+      "loss": 1.4124,
+      "step": 10334
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9730037450790405,
+      "learning_rate": 4.727068464181866e-05,
+      "loss": 1.3516,
+      "step": 10335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4300681352615356,
+      "learning_rate": 4.727042030134814e-05,
+      "loss": 1.3671,
+      "step": 10336
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5562875270843506,
+      "learning_rate": 4.727015596087761e-05,
+      "loss": 1.4144,
+      "step": 10337
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6355206966400146,
+      "learning_rate": 4.726989162040709e-05,
+      "loss": 1.3314,
+      "step": 10338
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6632907390594482,
+      "learning_rate": 4.726962727993656e-05,
+      "loss": 1.2882,
+      "step": 10339
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5451586246490479,
+      "learning_rate": 4.7269362939466035e-05,
+      "loss": 1.4765,
+      "step": 10340
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9293407201766968,
+      "learning_rate": 4.726909859899551e-05,
+      "loss": 1.6207,
+      "step": 10341
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6711055040359497,
+      "learning_rate": 4.726883425852498e-05,
+      "loss": 1.482,
+      "step": 10342
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7913743257522583,
+      "learning_rate": 4.7268569918054456e-05,
+      "loss": 1.4995,
+      "step": 10343
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0136330127716064,
+      "learning_rate": 4.726830557758393e-05,
+      "loss": 1.4395,
+      "step": 10344
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.322582483291626,
+      "learning_rate": 4.72680412371134e-05,
+      "loss": 1.194,
+      "step": 10345
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4534777402877808,
+      "learning_rate": 4.726777689664288e-05,
+      "loss": 1.4555,
+      "step": 10346
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9387527704238892,
+      "learning_rate": 4.726751255617235e-05,
+      "loss": 1.4258,
+      "step": 10347
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4832775592803955,
+      "learning_rate": 4.7267248215701824e-05,
+      "loss": 1.2644,
+      "step": 10348
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9079828262329102,
+      "learning_rate": 4.7266983875231304e-05,
+      "loss": 1.548,
+      "step": 10349
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7761571407318115,
+      "learning_rate": 4.726671953476077e-05,
+      "loss": 1.3206,
+      "step": 10350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0987322330474854,
+      "learning_rate": 4.726645519429025e-05,
+      "loss": 1.4591,
+      "step": 10351
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9838175773620605,
+      "learning_rate": 4.726619085381972e-05,
+      "loss": 1.4179,
+      "step": 10352
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6203917264938354,
+      "learning_rate": 4.72659265133492e-05,
+      "loss": 2.0155,
+      "step": 10353
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3761301040649414,
+      "learning_rate": 4.726566217287867e-05,
+      "loss": 1.503,
+      "step": 10354
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.785625696182251,
+      "learning_rate": 4.7265397832408146e-05,
+      "loss": 1.3748,
+      "step": 10355
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8632525205612183,
+      "learning_rate": 4.726513349193762e-05,
+      "loss": 1.4313,
+      "step": 10356
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.411301612854004,
+      "learning_rate": 4.7264869151467086e-05,
+      "loss": 1.5159,
+      "step": 10357
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6143813133239746,
+      "learning_rate": 4.7264604810996566e-05,
+      "loss": 1.4582,
+      "step": 10358
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.340659737586975,
+      "learning_rate": 4.726434047052604e-05,
+      "loss": 1.1993,
+      "step": 10359
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0367865562438965,
+      "learning_rate": 4.7264076130055514e-05,
+      "loss": 1.4844,
+      "step": 10360
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4610114097595215,
+      "learning_rate": 4.726381178958499e-05,
+      "loss": 1.3591,
+      "step": 10361
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4681891202926636,
+      "learning_rate": 4.726354744911446e-05,
+      "loss": 1.2577,
+      "step": 10362
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.611139178276062,
+      "learning_rate": 4.7263283108643934e-05,
+      "loss": 1.3402,
+      "step": 10363
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.038012981414795,
+      "learning_rate": 4.7263018768173415e-05,
+      "loss": 1.495,
+      "step": 10364
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7326282262802124,
+      "learning_rate": 4.726275442770288e-05,
+      "loss": 1.4379,
+      "step": 10365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1549487113952637,
+      "learning_rate": 4.7262490087232355e-05,
+      "loss": 1.6006,
+      "step": 10366
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6416783332824707,
+      "learning_rate": 4.726222574676183e-05,
+      "loss": 1.356,
+      "step": 10367
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5920584201812744,
+      "learning_rate": 4.72619614062913e-05,
+      "loss": 1.3711,
+      "step": 10368
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.080585479736328,
+      "learning_rate": 4.726169706582078e-05,
+      "loss": 1.5052,
+      "step": 10369
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.849198579788208,
+      "learning_rate": 4.726143272535025e-05,
+      "loss": 1.3329,
+      "step": 10370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.413307785987854,
+      "learning_rate": 4.726116838487973e-05,
+      "loss": 1.3416,
+      "step": 10371
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4791085720062256,
+      "learning_rate": 4.7260904044409196e-05,
+      "loss": 1.4501,
+      "step": 10372
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7183215618133545,
+      "learning_rate": 4.726063970393868e-05,
+      "loss": 1.5572,
+      "step": 10373
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0817606449127197,
+      "learning_rate": 4.726037536346815e-05,
+      "loss": 1.5888,
+      "step": 10374
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4084324836730957,
+      "learning_rate": 4.7260111022997624e-05,
+      "loss": 1.408,
+      "step": 10375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8532224893569946,
+      "learning_rate": 4.72598466825271e-05,
+      "loss": 1.4153,
+      "step": 10376
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7608070373535156,
+      "learning_rate": 4.725958234205657e-05,
+      "loss": 1.2904,
+      "step": 10377
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8861660957336426,
+      "learning_rate": 4.7259318001586045e-05,
+      "loss": 1.3624,
+      "step": 10378
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7890918254852295,
+      "learning_rate": 4.725905366111552e-05,
+      "loss": 1.298,
+      "step": 10379
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.011218786239624,
+      "learning_rate": 4.725878932064499e-05,
+      "loss": 1.4571,
+      "step": 10380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7614587545394897,
+      "learning_rate": 4.7258524980174465e-05,
+      "loss": 1.7591,
+      "step": 10381
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.042198657989502,
+      "learning_rate": 4.725826063970394e-05,
+      "loss": 1.4484,
+      "step": 10382
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.099357843399048,
+      "learning_rate": 4.725799629923341e-05,
+      "loss": 1.6006,
+      "step": 10383
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.705533742904663,
+      "learning_rate": 4.725773195876289e-05,
+      "loss": 1.439,
+      "step": 10384
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.610968828201294,
+      "learning_rate": 4.725746761829236e-05,
+      "loss": 1.8619,
+      "step": 10385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6479119062423706,
+      "learning_rate": 4.725720327782184e-05,
+      "loss": 1.4103,
+      "step": 10386
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9547609090805054,
+      "learning_rate": 4.725693893735131e-05,
+      "loss": 1.382,
+      "step": 10387
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9916877746582031,
+      "learning_rate": 4.725667459688079e-05,
+      "loss": 1.3949,
+      "step": 10388
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.684429407119751,
+      "learning_rate": 4.725641025641026e-05,
+      "loss": 1.1821,
+      "step": 10389
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.434355616569519,
+      "learning_rate": 4.7256145915939734e-05,
+      "loss": 1.3491,
+      "step": 10390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7714552879333496,
+      "learning_rate": 4.725588157546921e-05,
+      "loss": 1.3355,
+      "step": 10391
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8940246105194092,
+      "learning_rate": 4.725561723499868e-05,
+      "loss": 1.3424,
+      "step": 10392
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.059305191040039,
+      "learning_rate": 4.7255352894528155e-05,
+      "loss": 1.3975,
+      "step": 10393
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.988124132156372,
+      "learning_rate": 4.725508855405763e-05,
+      "loss": 1.706,
+      "step": 10394
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5790349245071411,
+      "learning_rate": 4.72548242135871e-05,
+      "loss": 1.2394,
+      "step": 10395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5800461769104004,
+      "learning_rate": 4.7254559873116576e-05,
+      "loss": 1.5711,
+      "step": 10396
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7277213335037231,
+      "learning_rate": 4.725429553264605e-05,
+      "loss": 1.4879,
+      "step": 10397
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5210000276565552,
+      "learning_rate": 4.725403119217552e-05,
+      "loss": 1.4346,
+      "step": 10398
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9318723678588867,
+      "learning_rate": 4.7253766851705e-05,
+      "loss": 1.2731,
+      "step": 10399
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6823582649230957,
+      "learning_rate": 4.725350251123447e-05,
+      "loss": 1.4344,
+      "step": 10400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.040285110473633,
+      "learning_rate": 4.725323817076395e-05,
+      "loss": 1.6927,
+      "step": 10401
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.385987401008606,
+      "learning_rate": 4.725297383029342e-05,
+      "loss": 1.7418,
+      "step": 10402
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0457634925842285,
+      "learning_rate": 4.725270948982289e-05,
+      "loss": 1.5994,
+      "step": 10403
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.823625922203064,
+      "learning_rate": 4.725244514935237e-05,
+      "loss": 1.4912,
+      "step": 10404
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2277402877807617,
+      "learning_rate": 4.725218080888184e-05,
+      "loss": 1.7522,
+      "step": 10405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.777546763420105,
+      "learning_rate": 4.725191646841132e-05,
+      "loss": 1.3201,
+      "step": 10406
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.843498706817627,
+      "learning_rate": 4.7251652127940785e-05,
+      "loss": 1.4347,
+      "step": 10407
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9018523693084717,
+      "learning_rate": 4.7251387787470265e-05,
+      "loss": 1.4419,
+      "step": 10408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8546826839447021,
+      "learning_rate": 4.725112344699974e-05,
+      "loss": 1.3518,
+      "step": 10409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.711987018585205,
+      "learning_rate": 4.725085910652921e-05,
+      "loss": 1.5622,
+      "step": 10410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8972824811935425,
+      "learning_rate": 4.7250594766058686e-05,
+      "loss": 1.5239,
+      "step": 10411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0536153316497803,
+      "learning_rate": 4.725033042558816e-05,
+      "loss": 1.5066,
+      "step": 10412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.649644374847412,
+      "learning_rate": 4.725006608511763e-05,
+      "loss": 1.3994,
+      "step": 10413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0802078247070312,
+      "learning_rate": 4.724980174464711e-05,
+      "loss": 1.4594,
+      "step": 10414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.842624545097351,
+      "learning_rate": 4.724953740417658e-05,
+      "loss": 1.3642,
+      "step": 10415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8490064144134521,
+      "learning_rate": 4.7249273063706054e-05,
+      "loss": 1.3714,
+      "step": 10416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8691829442977905,
+      "learning_rate": 4.724900872323553e-05,
+      "loss": 1.4676,
+      "step": 10417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7713655233383179,
+      "learning_rate": 4.7248744382765e-05,
+      "loss": 1.2577,
+      "step": 10418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8015910387039185,
+      "learning_rate": 4.724848004229448e-05,
+      "loss": 1.2643,
+      "step": 10419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7402969598770142,
+      "learning_rate": 4.724821570182395e-05,
+      "loss": 1.3758,
+      "step": 10420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.430914282798767,
+      "learning_rate": 4.724795136135343e-05,
+      "loss": 1.6287,
+      "step": 10421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6496803760528564,
+      "learning_rate": 4.7247687020882895e-05,
+      "loss": 1.423,
+      "step": 10422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4878952503204346,
+      "learning_rate": 4.7247422680412376e-05,
+      "loss": 1.2912,
+      "step": 10423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5366637706756592,
+      "learning_rate": 4.724715833994185e-05,
+      "loss": 1.942,
+      "step": 10424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8953498601913452,
+      "learning_rate": 4.724689399947132e-05,
+      "loss": 1.4362,
+      "step": 10425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4987456798553467,
+      "learning_rate": 4.7246629659000796e-05,
+      "loss": 1.476,
+      "step": 10426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9478161334991455,
+      "learning_rate": 4.724636531853027e-05,
+      "loss": 1.5301,
+      "step": 10427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4332245588302612,
+      "learning_rate": 4.7246100978059744e-05,
+      "loss": 1.3269,
+      "step": 10428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8735970258712769,
+      "learning_rate": 4.724583663758922e-05,
+      "loss": 1.4468,
+      "step": 10429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.815070629119873,
+      "learning_rate": 4.724557229711869e-05,
+      "loss": 1.4562,
+      "step": 10430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6744327545166016,
+      "learning_rate": 4.7245307956648164e-05,
+      "loss": 1.6649,
+      "step": 10431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.481422781944275,
+      "learning_rate": 4.724504361617764e-05,
+      "loss": 2.2975,
+      "step": 10432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7293968200683594,
+      "learning_rate": 4.724477927570711e-05,
+      "loss": 1.4723,
+      "step": 10433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4554951190948486,
+      "learning_rate": 4.724451493523659e-05,
+      "loss": 1.5822,
+      "step": 10434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4799455404281616,
+      "learning_rate": 4.724425059476606e-05,
+      "loss": 1.2871,
+      "step": 10435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9609787464141846,
+      "learning_rate": 4.724398625429554e-05,
+      "loss": 1.4936,
+      "step": 10436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6719926595687866,
+      "learning_rate": 4.7243721913825006e-05,
+      "loss": 1.3508,
+      "step": 10437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6265387535095215,
+      "learning_rate": 4.724345757335448e-05,
+      "loss": 1.3605,
+      "step": 10438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0474231243133545,
+      "learning_rate": 4.724319323288396e-05,
+      "loss": 1.4178,
+      "step": 10439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6522506475448608,
+      "learning_rate": 4.7242928892413426e-05,
+      "loss": 1.4243,
+      "step": 10440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.881770610809326,
+      "learning_rate": 4.724266455194291e-05,
+      "loss": 1.5535,
+      "step": 10441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5621473789215088,
+      "learning_rate": 4.7242400211472374e-05,
+      "loss": 1.3046,
+      "step": 10442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8973057270050049,
+      "learning_rate": 4.7242135871001854e-05,
+      "loss": 1.4331,
+      "step": 10443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1130239963531494,
+      "learning_rate": 4.724187153053133e-05,
+      "loss": 1.4289,
+      "step": 10444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7606326341629028,
+      "learning_rate": 4.72416071900608e-05,
+      "loss": 1.4572,
+      "step": 10445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8955987691879272,
+      "learning_rate": 4.7241342849590275e-05,
+      "loss": 1.331,
+      "step": 10446
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6357917785644531,
+      "learning_rate": 4.724107850911975e-05,
+      "loss": 1.3582,
+      "step": 10447
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.005185604095459,
+      "learning_rate": 4.724081416864922e-05,
+      "loss": 1.4121,
+      "step": 10448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7401695251464844,
+      "learning_rate": 4.7240549828178695e-05,
+      "loss": 1.292,
+      "step": 10449
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2853076457977295,
+      "learning_rate": 4.724028548770817e-05,
+      "loss": 1.4101,
+      "step": 10450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3875112533569336,
+      "learning_rate": 4.724002114723764e-05,
+      "loss": 1.4115,
+      "step": 10451
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0675432682037354,
+      "learning_rate": 4.7239756806767116e-05,
+      "loss": 1.451,
+      "step": 10452
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.434832215309143,
+      "learning_rate": 4.723949246629659e-05,
+      "loss": 1.2504,
+      "step": 10453
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.598543643951416,
+      "learning_rate": 4.723922812582607e-05,
+      "loss": 1.3026,
+      "step": 10454
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4360620975494385,
+      "learning_rate": 4.723896378535554e-05,
+      "loss": 1.3445,
+      "step": 10455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.595287561416626,
+      "learning_rate": 4.723869944488502e-05,
+      "loss": 1.2287,
+      "step": 10456
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.909889578819275,
+      "learning_rate": 4.7238435104414484e-05,
+      "loss": 1.347,
+      "step": 10457
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6997779607772827,
+      "learning_rate": 4.7238170763943964e-05,
+      "loss": 1.2753,
+      "step": 10458
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.037742853164673,
+      "learning_rate": 4.723790642347344e-05,
+      "loss": 1.5387,
+      "step": 10459
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.205049514770508,
+      "learning_rate": 4.723764208300291e-05,
+      "loss": 1.7286,
+      "step": 10460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7638006210327148,
+      "learning_rate": 4.7237377742532385e-05,
+      "loss": 1.2861,
+      "step": 10461
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.146853446960449,
+      "learning_rate": 4.723711340206186e-05,
+      "loss": 1.4082,
+      "step": 10462
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3537778854370117,
+      "learning_rate": 4.723684906159133e-05,
+      "loss": 1.5004,
+      "step": 10463
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.43303108215332,
+      "learning_rate": 4.7236584721120806e-05,
+      "loss": 1.5611,
+      "step": 10464
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9652080535888672,
+      "learning_rate": 4.723632038065028e-05,
+      "loss": 1.5147,
+      "step": 10465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.255237102508545,
+      "learning_rate": 4.723605604017975e-05,
+      "loss": 1.5759,
+      "step": 10466
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6234569549560547,
+      "learning_rate": 4.7235791699709226e-05,
+      "loss": 1.5387,
+      "step": 10467
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1147167682647705,
+      "learning_rate": 4.72355273592387e-05,
+      "loss": 1.3165,
+      "step": 10468
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.008523941040039,
+      "learning_rate": 4.723526301876818e-05,
+      "loss": 1.4712,
+      "step": 10469
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.104443073272705,
+      "learning_rate": 4.723499867829765e-05,
+      "loss": 1.4218,
+      "step": 10470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.755852460861206,
+      "learning_rate": 4.723473433782713e-05,
+      "loss": 1.3445,
+      "step": 10471
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.23852801322937,
+      "learning_rate": 4.7234469997356594e-05,
+      "loss": 1.4405,
+      "step": 10472
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9289147853851318,
+      "learning_rate": 4.7234205656886075e-05,
+      "loss": 1.4617,
+      "step": 10473
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4923458099365234,
+      "learning_rate": 4.723394131641555e-05,
+      "loss": 1.4675,
+      "step": 10474
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.177518606185913,
+      "learning_rate": 4.7233676975945015e-05,
+      "loss": 1.3905,
+      "step": 10475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5128439664840698,
+      "learning_rate": 4.7233412635474495e-05,
+      "loss": 1.2672,
+      "step": 10476
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7361688613891602,
+      "learning_rate": 4.723314829500396e-05,
+      "loss": 1.4625,
+      "step": 10477
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.609694480895996,
+      "learning_rate": 4.723288395453344e-05,
+      "loss": 1.598,
+      "step": 10478
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1145846843719482,
+      "learning_rate": 4.7232619614062916e-05,
+      "loss": 1.3742,
+      "step": 10479
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9721118211746216,
+      "learning_rate": 4.723235527359239e-05,
+      "loss": 1.4368,
+      "step": 10480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0304481983184814,
+      "learning_rate": 4.723209093312186e-05,
+      "loss": 1.7412,
+      "step": 10481
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7003072500228882,
+      "learning_rate": 4.723182659265134e-05,
+      "loss": 1.2168,
+      "step": 10482
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3021469116210938,
+      "learning_rate": 4.723156225218081e-05,
+      "loss": 1.554,
+      "step": 10483
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8353183269500732,
+      "learning_rate": 4.7231297911710284e-05,
+      "loss": 1.4381,
+      "step": 10484
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5573242902755737,
+      "learning_rate": 4.723103357123976e-05,
+      "loss": 1.3071,
+      "step": 10485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2742831707000732,
+      "learning_rate": 4.723076923076923e-05,
+      "loss": 1.4893,
+      "step": 10486
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7363559007644653,
+      "learning_rate": 4.7230504890298705e-05,
+      "loss": 1.3646,
+      "step": 10487
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7507380247116089,
+      "learning_rate": 4.723024054982818e-05,
+      "loss": 1.356,
+      "step": 10488
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4950791597366333,
+      "learning_rate": 4.722997620935766e-05,
+      "loss": 1.2227,
+      "step": 10489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.419355034828186,
+      "learning_rate": 4.7229711868887125e-05,
+      "loss": 1.3213,
+      "step": 10490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3424402475357056,
+      "learning_rate": 4.7229447528416606e-05,
+      "loss": 1.1434,
+      "step": 10491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3197312355041504,
+      "learning_rate": 4.722918318794607e-05,
+      "loss": 1.3655,
+      "step": 10492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3981884717941284,
+      "learning_rate": 4.722891884747555e-05,
+      "loss": 1.625,
+      "step": 10493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3320724964141846,
+      "learning_rate": 4.7228654507005026e-05,
+      "loss": 1.4802,
+      "step": 10494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.262551784515381,
+      "learning_rate": 4.72283901665345e-05,
+      "loss": 1.6101,
+      "step": 10495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8794536590576172,
+      "learning_rate": 4.7228125826063973e-05,
+      "loss": 1.3363,
+      "step": 10496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.975656509399414,
+      "learning_rate": 4.722786148559345e-05,
+      "loss": 1.4646,
+      "step": 10497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4643967151641846,
+      "learning_rate": 4.722759714512292e-05,
+      "loss": 1.3911,
+      "step": 10498
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7242461442947388,
+      "learning_rate": 4.7227332804652394e-05,
+      "loss": 1.3471,
+      "step": 10499
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.279071569442749,
+      "learning_rate": 4.722706846418187e-05,
+      "loss": 1.9856,
+      "step": 10500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7792719602584839,
+      "learning_rate": 4.722680412371134e-05,
+      "loss": 1.4717,
+      "step": 10501
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.194648265838623,
+      "learning_rate": 4.7226539783240815e-05,
+      "loss": 2.6678,
+      "step": 10502
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1485466957092285,
+      "learning_rate": 4.722627544277029e-05,
+      "loss": 1.4882,
+      "step": 10503
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0745134353637695,
+      "learning_rate": 4.722601110229977e-05,
+      "loss": 1.4126,
+      "step": 10504
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.963120698928833,
+      "learning_rate": 4.7225746761829236e-05,
+      "loss": 1.5567,
+      "step": 10505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.786710500717163,
+      "learning_rate": 4.7225482421358716e-05,
+      "loss": 1.4274,
+      "step": 10506
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6730743646621704,
+      "learning_rate": 4.722521808088818e-05,
+      "loss": 1.4048,
+      "step": 10507
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0732760429382324,
+      "learning_rate": 4.722495374041766e-05,
+      "loss": 1.5374,
+      "step": 10508
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8942149877548218,
+      "learning_rate": 4.722468939994714e-05,
+      "loss": 1.488,
+      "step": 10509
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5635021924972534,
+      "learning_rate": 4.722442505947661e-05,
+      "loss": 1.3076,
+      "step": 10510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.155803680419922,
+      "learning_rate": 4.7224160719006084e-05,
+      "loss": 1.4422,
+      "step": 10511
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0436089038848877,
+      "learning_rate": 4.722389637853555e-05,
+      "loss": 1.4946,
+      "step": 10512
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2801616191864014,
+      "learning_rate": 4.722363203806503e-05,
+      "loss": 1.323,
+      "step": 10513
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.88027822971344,
+      "learning_rate": 4.7223367697594505e-05,
+      "loss": 1.3404,
+      "step": 10514
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6607880592346191,
+      "learning_rate": 4.722310335712398e-05,
+      "loss": 1.4023,
+      "step": 10515
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.178797960281372,
+      "learning_rate": 4.722283901665345e-05,
+      "loss": 1.6268,
+      "step": 10516
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3660389184951782,
+      "learning_rate": 4.7222574676182925e-05,
+      "loss": 1.3116,
+      "step": 10517
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5958905220031738,
+      "learning_rate": 4.72223103357124e-05,
+      "loss": 1.5972,
+      "step": 10518
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4943667650222778,
+      "learning_rate": 4.722204599524188e-05,
+      "loss": 1.4153,
+      "step": 10519
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7672240734100342,
+      "learning_rate": 4.7221781654771346e-05,
+      "loss": 1.4804,
+      "step": 10520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.962293267250061,
+      "learning_rate": 4.722151731430082e-05,
+      "loss": 1.4632,
+      "step": 10521
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5667579174041748,
+      "learning_rate": 4.722125297383029e-05,
+      "loss": 1.2299,
+      "step": 10522
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4328043460845947,
+      "learning_rate": 4.722098863335977e-05,
+      "loss": 1.4069,
+      "step": 10523
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7904627323150635,
+      "learning_rate": 4.722072429288925e-05,
+      "loss": 1.3554,
+      "step": 10524
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.146451950073242,
+      "learning_rate": 4.7220459952418714e-05,
+      "loss": 1.3275,
+      "step": 10525
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.061398506164551,
+      "learning_rate": 4.7220195611948194e-05,
+      "loss": 1.5708,
+      "step": 10526
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.890200138092041,
+      "learning_rate": 4.721993127147766e-05,
+      "loss": 1.4316,
+      "step": 10527
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5276521444320679,
+      "learning_rate": 4.721966693100714e-05,
+      "loss": 1.4127,
+      "step": 10528
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.938101887702942,
+      "learning_rate": 4.7219402590536615e-05,
+      "loss": 1.6189,
+      "step": 10529
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.851658821105957,
+      "learning_rate": 4.721913825006609e-05,
+      "loss": 1.4581,
+      "step": 10530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.199367046356201,
+      "learning_rate": 4.721887390959556e-05,
+      "loss": 1.4848,
+      "step": 10531
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8560963869094849,
+      "learning_rate": 4.7218609569125036e-05,
+      "loss": 1.4618,
+      "step": 10532
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.593605637550354,
+      "learning_rate": 4.721834522865451e-05,
+      "loss": 1.3635,
+      "step": 10533
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7638195753097534,
+      "learning_rate": 4.721808088818398e-05,
+      "loss": 1.3288,
+      "step": 10534
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6970813274383545,
+      "learning_rate": 4.7217816547713456e-05,
+      "loss": 1.4581,
+      "step": 10535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.584083914756775,
+      "learning_rate": 4.721755220724293e-05,
+      "loss": 1.3285,
+      "step": 10536
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.071403980255127,
+      "learning_rate": 4.7217287866772403e-05,
+      "loss": 1.5441,
+      "step": 10537
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.528961420059204,
+      "learning_rate": 4.721702352630188e-05,
+      "loss": 1.408,
+      "step": 10538
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9083385467529297,
+      "learning_rate": 4.721675918583136e-05,
+      "loss": 1.4724,
+      "step": 10539
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.971604347229004,
+      "learning_rate": 4.7216494845360824e-05,
+      "loss": 1.5062,
+      "step": 10540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984247088432312,
+      "learning_rate": 4.7216230504890305e-05,
+      "loss": 1.4231,
+      "step": 10541
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.963822364807129,
+      "learning_rate": 4.721596616441977e-05,
+      "loss": 1.4532,
+      "step": 10542
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6032789945602417,
+      "learning_rate": 4.721570182394925e-05,
+      "loss": 1.344,
+      "step": 10543
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6333378553390503,
+      "learning_rate": 4.7215437483478725e-05,
+      "loss": 1.7413,
+      "step": 10544
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9539109468460083,
+      "learning_rate": 4.72151731430082e-05,
+      "loss": 1.61,
+      "step": 10545
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.341131567955017,
+      "learning_rate": 4.721490880253767e-05,
+      "loss": 1.2602,
+      "step": 10546
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.217620849609375,
+      "learning_rate": 4.721464446206714e-05,
+      "loss": 1.2395,
+      "step": 10547
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7982738018035889,
+      "learning_rate": 4.721438012159662e-05,
+      "loss": 1.4969,
+      "step": 10548
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7865453958511353,
+      "learning_rate": 4.721411578112609e-05,
+      "loss": 1.3997,
+      "step": 10549
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6692792177200317,
+      "learning_rate": 4.721385144065557e-05,
+      "loss": 2.1311,
+      "step": 10550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5950746536254883,
+      "learning_rate": 4.721358710018504e-05,
+      "loss": 1.2311,
+      "step": 10551
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7107303142547607,
+      "learning_rate": 4.7213322759714514e-05,
+      "loss": 1.2495,
+      "step": 10552
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6365692615509033,
+      "learning_rate": 4.721305841924399e-05,
+      "loss": 1.3178,
+      "step": 10553
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3186018466949463,
+      "learning_rate": 4.721279407877347e-05,
+      "loss": 1.594,
+      "step": 10554
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5046945810317993,
+      "learning_rate": 4.7212529738302935e-05,
+      "loss": 1.2892,
+      "step": 10555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7706573009490967,
+      "learning_rate": 4.721226539783241e-05,
+      "loss": 1.4723,
+      "step": 10556
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6051164865493774,
+      "learning_rate": 4.721200105736188e-05,
+      "loss": 2.0575,
+      "step": 10557
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7500874996185303,
+      "learning_rate": 4.7211736716891355e-05,
+      "loss": 1.3952,
+      "step": 10558
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1458466053009033,
+      "learning_rate": 4.7211472376420836e-05,
+      "loss": 1.3286,
+      "step": 10559
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8622432947158813,
+      "learning_rate": 4.72112080359503e-05,
+      "loss": 1.3099,
+      "step": 10560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9269965887069702,
+      "learning_rate": 4.721094369547978e-05,
+      "loss": 1.3371,
+      "step": 10561
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.922748327255249,
+      "learning_rate": 4.721067935500925e-05,
+      "loss": 1.4499,
+      "step": 10562
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.123455286026001,
+      "learning_rate": 4.721041501453873e-05,
+      "loss": 1.3711,
+      "step": 10563
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.504333734512329,
+      "learning_rate": 4.7210150674068203e-05,
+      "loss": 1.3455,
+      "step": 10564
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.707721471786499,
+      "learning_rate": 4.720988633359768e-05,
+      "loss": 1.5007,
+      "step": 10565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6526917219161987,
+      "learning_rate": 4.720962199312715e-05,
+      "loss": 1.4374,
+      "step": 10566
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.07731294631958,
+      "learning_rate": 4.7209357652656624e-05,
+      "loss": 1.319,
+      "step": 10567
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.219475269317627,
+      "learning_rate": 4.72090933121861e-05,
+      "loss": 1.2455,
+      "step": 10568
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0194685459136963,
+      "learning_rate": 4.720882897171557e-05,
+      "loss": 1.3689,
+      "step": 10569
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6142356395721436,
+      "learning_rate": 4.7208564631245045e-05,
+      "loss": 1.5251,
+      "step": 10570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8894587755203247,
+      "learning_rate": 4.720830029077452e-05,
+      "loss": 1.4075,
+      "step": 10571
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7200790643692017,
+      "learning_rate": 4.720803595030399e-05,
+      "loss": 1.3106,
+      "step": 10572
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.863053321838379,
+      "learning_rate": 4.7207771609833466e-05,
+      "loss": 1.5257,
+      "step": 10573
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7326281070709229,
+      "learning_rate": 4.7207507269362946e-05,
+      "loss": 1.4427,
+      "step": 10574
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1972386837005615,
+      "learning_rate": 4.720724292889241e-05,
+      "loss": 1.3587,
+      "step": 10575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1267313957214355,
+      "learning_rate": 4.720697858842189e-05,
+      "loss": 1.4062,
+      "step": 10576
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5272576808929443,
+      "learning_rate": 4.720671424795136e-05,
+      "loss": 1.2825,
+      "step": 10577
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7858659029006958,
+      "learning_rate": 4.720644990748084e-05,
+      "loss": 1.5822,
+      "step": 10578
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.19033145904541,
+      "learning_rate": 4.7206185567010314e-05,
+      "loss": 1.5081,
+      "step": 10579
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7635729312896729,
+      "learning_rate": 4.720592122653979e-05,
+      "loss": 2.4406,
+      "step": 10580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.519808292388916,
+      "learning_rate": 4.720565688606926e-05,
+      "loss": 1.4159,
+      "step": 10581
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6426634788513184,
+      "learning_rate": 4.7205392545598734e-05,
+      "loss": 1.3886,
+      "step": 10582
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7268776893615723,
+      "learning_rate": 4.720512820512821e-05,
+      "loss": 1.4331,
+      "step": 10583
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4250004291534424,
+      "learning_rate": 4.720486386465768e-05,
+      "loss": 1.4946,
+      "step": 10584
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.857149600982666,
+      "learning_rate": 4.7204599524187155e-05,
+      "loss": 1.5206,
+      "step": 10585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6633683443069458,
+      "learning_rate": 4.720433518371663e-05,
+      "loss": 1.3141,
+      "step": 10586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0498719215393066,
+      "learning_rate": 4.72040708432461e-05,
+      "loss": 1.4518,
+      "step": 10587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4291412830352783,
+      "learning_rate": 4.7203806502775576e-05,
+      "loss": 1.4093,
+      "step": 10588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5796836614608765,
+      "learning_rate": 4.7203542162305056e-05,
+      "loss": 1.2673,
+      "step": 10589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4147624969482422,
+      "learning_rate": 4.720327782183452e-05,
+      "loss": 1.1752,
+      "step": 10590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.700459361076355,
+      "learning_rate": 4.7203013481364003e-05,
+      "loss": 1.328,
+      "step": 10591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4770731925964355,
+      "learning_rate": 4.720274914089347e-05,
+      "loss": 1.37,
+      "step": 10592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1277735233306885,
+      "learning_rate": 4.7202484800422944e-05,
+      "loss": 1.4183,
+      "step": 10593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.08211612701416,
+      "learning_rate": 4.7202220459952424e-05,
+      "loss": 1.2776,
+      "step": 10594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6814987659454346,
+      "learning_rate": 4.720195611948189e-05,
+      "loss": 1.4357,
+      "step": 10595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2997963428497314,
+      "learning_rate": 4.720169177901137e-05,
+      "loss": 1.4813,
+      "step": 10596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.593612551689148,
+      "learning_rate": 4.720142743854084e-05,
+      "loss": 1.3246,
+      "step": 10597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7548636198043823,
+      "learning_rate": 4.720116309807032e-05,
+      "loss": 1.5319,
+      "step": 10598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7851780652999878,
+      "learning_rate": 4.720089875759979e-05,
+      "loss": 1.5041,
+      "step": 10599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.991371989250183,
+      "learning_rate": 4.7200634417129266e-05,
+      "loss": 1.305,
+      "step": 10600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3572173118591309,
+      "learning_rate": 4.720037007665874e-05,
+      "loss": 1.3691,
+      "step": 10601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5274338722229004,
+      "learning_rate": 4.720010573618821e-05,
+      "loss": 1.1675,
+      "step": 10602
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.522844672203064,
+      "learning_rate": 4.7199841395717686e-05,
+      "loss": 1.3259,
+      "step": 10603
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2443182468414307,
+      "learning_rate": 4.719957705524716e-05,
+      "loss": 1.461,
+      "step": 10604
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7058569192886353,
+      "learning_rate": 4.719931271477663e-05,
+      "loss": 1.4308,
+      "step": 10605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9234085083007812,
+      "learning_rate": 4.719904837430611e-05,
+      "loss": 1.3871,
+      "step": 10606
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8353533744812012,
+      "learning_rate": 4.719878403383558e-05,
+      "loss": 1.4006,
+      "step": 10607
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.844074010848999,
+      "learning_rate": 4.7198519693365054e-05,
+      "loss": 1.3291,
+      "step": 10608
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9485856294631958,
+      "learning_rate": 4.7198255352894534e-05,
+      "loss": 1.3963,
+      "step": 10609
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8666349649429321,
+      "learning_rate": 4.7197991012424e-05,
+      "loss": 1.2697,
+      "step": 10610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8053011894226074,
+      "learning_rate": 4.719772667195348e-05,
+      "loss": 1.523,
+      "step": 10611
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0143094062805176,
+      "learning_rate": 4.719746233148295e-05,
+      "loss": 1.4042,
+      "step": 10612
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7319782972335815,
+      "learning_rate": 4.719719799101243e-05,
+      "loss": 1.5652,
+      "step": 10613
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.683803915977478,
+      "learning_rate": 4.71969336505419e-05,
+      "loss": 1.4498,
+      "step": 10614
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7244950532913208,
+      "learning_rate": 4.7196669310071376e-05,
+      "loss": 1.3962,
+      "step": 10615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7586743831634521,
+      "learning_rate": 4.719640496960085e-05,
+      "loss": 1.4974,
+      "step": 10616
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5923629999160767,
+      "learning_rate": 4.719614062913032e-05,
+      "loss": 1.2738,
+      "step": 10617
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6419264078140259,
+      "learning_rate": 4.7195876288659797e-05,
+      "loss": 1.3269,
+      "step": 10618
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1262283325195312,
+      "learning_rate": 4.719561194818927e-05,
+      "loss": 1.3982,
+      "step": 10619
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8410471677780151,
+      "learning_rate": 4.7195347607718744e-05,
+      "loss": 1.5604,
+      "step": 10620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6617305278778076,
+      "learning_rate": 4.719508326724822e-05,
+      "loss": 1.3654,
+      "step": 10621
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0330898761749268,
+      "learning_rate": 4.719481892677769e-05,
+      "loss": 2.734,
+      "step": 10622
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2168495655059814,
+      "learning_rate": 4.7194554586307164e-05,
+      "loss": 1.302,
+      "step": 10623
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0322861671447754,
+      "learning_rate": 4.7194290245836645e-05,
+      "loss": 1.4985,
+      "step": 10624
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5773581266403198,
+      "learning_rate": 4.719402590536611e-05,
+      "loss": 1.2355,
+      "step": 10625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7921030521392822,
+      "learning_rate": 4.719376156489559e-05,
+      "loss": 1.436,
+      "step": 10626
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.039008140563965,
+      "learning_rate": 4.719349722442506e-05,
+      "loss": 1.5345,
+      "step": 10627
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9669735431671143,
+      "learning_rate": 4.719323288395454e-05,
+      "loss": 1.5511,
+      "step": 10628
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0823042392730713,
+      "learning_rate": 4.719296854348401e-05,
+      "loss": 1.3631,
+      "step": 10629
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3667616844177246,
+      "learning_rate": 4.719270420301348e-05,
+      "loss": 1.3004,
+      "step": 10630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8986279964447021,
+      "learning_rate": 4.719243986254296e-05,
+      "loss": 1.3926,
+      "step": 10631
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6265807151794434,
+      "learning_rate": 4.7192175522072427e-05,
+      "loss": 1.2593,
+      "step": 10632
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5615546703338623,
+      "learning_rate": 4.719191118160191e-05,
+      "loss": 2.1793,
+      "step": 10633
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7546838521957397,
+      "learning_rate": 4.719164684113138e-05,
+      "loss": 1.3753,
+      "step": 10634
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6737216711044312,
+      "learning_rate": 4.7191382500660854e-05,
+      "loss": 1.2513,
+      "step": 10635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.952644944190979,
+      "learning_rate": 4.719111816019033e-05,
+      "loss": 1.3989,
+      "step": 10636
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.423832893371582,
+      "learning_rate": 4.71908538197198e-05,
+      "loss": 1.6307,
+      "step": 10637
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1020424365997314,
+      "learning_rate": 4.7190589479249275e-05,
+      "loss": 1.5021,
+      "step": 10638
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4712281227111816,
+      "learning_rate": 4.719032513877875e-05,
+      "loss": 1.3877,
+      "step": 10639
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8423492908477783,
+      "learning_rate": 4.719006079830822e-05,
+      "loss": 1.256,
+      "step": 10640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.561326026916504,
+      "learning_rate": 4.7189796457837695e-05,
+      "loss": 1.3626,
+      "step": 10641
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6006999015808105,
+      "learning_rate": 4.718953211736717e-05,
+      "loss": 1.3793,
+      "step": 10642
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.923699975013733,
+      "learning_rate": 4.718926777689664e-05,
+      "loss": 1.4883,
+      "step": 10643
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5617527961730957,
+      "learning_rate": 4.718900343642612e-05,
+      "loss": 1.3244,
+      "step": 10644
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.081914186477661,
+      "learning_rate": 4.718873909595559e-05,
+      "loss": 1.6233,
+      "step": 10645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7943464517593384,
+      "learning_rate": 4.718847475548507e-05,
+      "loss": 1.3406,
+      "step": 10646
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9169317483901978,
+      "learning_rate": 4.718821041501454e-05,
+      "loss": 1.4014,
+      "step": 10647
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8686718940734863,
+      "learning_rate": 4.718794607454402e-05,
+      "loss": 1.6142,
+      "step": 10648
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8315941095352173,
+      "learning_rate": 4.718768173407349e-05,
+      "loss": 1.3308,
+      "step": 10649
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7649871110916138,
+      "learning_rate": 4.7187417393602964e-05,
+      "loss": 1.4825,
+      "step": 10650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2054600715637207,
+      "learning_rate": 4.718715305313244e-05,
+      "loss": 1.4823,
+      "step": 10651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4809789657592773,
+      "learning_rate": 4.718688871266191e-05,
+      "loss": 1.3649,
+      "step": 10652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3639779090881348,
+      "learning_rate": 4.7186624372191385e-05,
+      "loss": 1.7668,
+      "step": 10653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.030877113342285,
+      "learning_rate": 4.718636003172086e-05,
+      "loss": 1.5072,
+      "step": 10654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4079902172088623,
+      "learning_rate": 4.718609569125033e-05,
+      "loss": 1.5329,
+      "step": 10655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.246954917907715,
+      "learning_rate": 4.7185831350779806e-05,
+      "loss": 1.4694,
+      "step": 10656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6635820865631104,
+      "learning_rate": 4.718556701030928e-05,
+      "loss": 1.5036,
+      "step": 10657
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.215392589569092,
+      "learning_rate": 4.718530266983875e-05,
+      "loss": 1.5267,
+      "step": 10658
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5073875188827515,
+      "learning_rate": 4.718503832936823e-05,
+      "loss": 1.2353,
+      "step": 10659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.450336217880249,
+      "learning_rate": 4.71847739888977e-05,
+      "loss": 1.2848,
+      "step": 10660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8115246295928955,
+      "learning_rate": 4.718450964842718e-05,
+      "loss": 1.4983,
+      "step": 10661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8950538635253906,
+      "learning_rate": 4.718424530795665e-05,
+      "loss": 1.4195,
+      "step": 10662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.051042079925537,
+      "learning_rate": 4.718398096748613e-05,
+      "loss": 1.5104,
+      "step": 10663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9443912506103516,
+      "learning_rate": 4.71837166270156e-05,
+      "loss": 2.7469,
+      "step": 10664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9837720394134521,
+      "learning_rate": 4.718345228654507e-05,
+      "loss": 1.5144,
+      "step": 10665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6236475706100464,
+      "learning_rate": 4.718318794607455e-05,
+      "loss": 1.2271,
+      "step": 10666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4199633598327637,
+      "learning_rate": 4.7182923605604015e-05,
+      "loss": 1.4113,
+      "step": 10667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9909238815307617,
+      "learning_rate": 4.7182659265133495e-05,
+      "loss": 1.2686,
+      "step": 10668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.501255512237549,
+      "learning_rate": 4.718239492466297e-05,
+      "loss": 1.3944,
+      "step": 10669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.185758590698242,
+      "learning_rate": 4.718213058419244e-05,
+      "loss": 1.5667,
+      "step": 10670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7836973667144775,
+      "learning_rate": 4.7181866243721916e-05,
+      "loss": 1.3751,
+      "step": 10671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.356421709060669,
+      "learning_rate": 4.718160190325139e-05,
+      "loss": 1.6034,
+      "step": 10672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.64565110206604,
+      "learning_rate": 4.718133756278086e-05,
+      "loss": 1.3551,
+      "step": 10673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5012447834014893,
+      "learning_rate": 4.718107322231034e-05,
+      "loss": 1.4217,
+      "step": 10674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.645041823387146,
+      "learning_rate": 4.718080888183981e-05,
+      "loss": 1.4288,
+      "step": 10675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9065821170806885,
+      "learning_rate": 4.7180544541369284e-05,
+      "loss": 1.2492,
+      "step": 10676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.242157459259033,
+      "learning_rate": 4.718028020089876e-05,
+      "loss": 1.5045,
+      "step": 10677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8143987655639648,
+      "learning_rate": 4.718001586042823e-05,
+      "loss": 1.1898,
+      "step": 10678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.738133192062378,
+      "learning_rate": 4.717975151995771e-05,
+      "loss": 2.1551,
+      "step": 10679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.250253438949585,
+      "learning_rate": 4.717948717948718e-05,
+      "loss": 1.4884,
+      "step": 10680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6026054620742798,
+      "learning_rate": 4.717922283901666e-05,
+      "loss": 1.3641,
+      "step": 10681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2383291721343994,
+      "learning_rate": 4.7178958498546125e-05,
+      "loss": 1.5408,
+      "step": 10682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5683132410049438,
+      "learning_rate": 4.7178694158075606e-05,
+      "loss": 1.3604,
+      "step": 10683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6355503797531128,
+      "learning_rate": 4.717842981760508e-05,
+      "loss": 1.2651,
+      "step": 10684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8123459815979004,
+      "learning_rate": 4.717816547713455e-05,
+      "loss": 1.5518,
+      "step": 10685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.076756238937378,
+      "learning_rate": 4.7177901136664027e-05,
+      "loss": 1.6254,
+      "step": 10686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.653074026107788,
+      "learning_rate": 4.71776367961935e-05,
+      "loss": 1.3622,
+      "step": 10687
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7624855041503906,
+      "learning_rate": 4.7177372455722974e-05,
+      "loss": 1.4326,
+      "step": 10688
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.686824083328247,
+      "learning_rate": 4.717710811525245e-05,
+      "loss": 1.4663,
+      "step": 10689
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.746317982673645,
+      "learning_rate": 4.717684377478192e-05,
+      "loss": 1.3925,
+      "step": 10690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1097412109375,
+      "learning_rate": 4.7176579434311394e-05,
+      "loss": 1.4494,
+      "step": 10691
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7670741081237793,
+      "learning_rate": 4.717631509384087e-05,
+      "loss": 1.3756,
+      "step": 10692
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.939570426940918,
+      "learning_rate": 4.717605075337034e-05,
+      "loss": 1.3414,
+      "step": 10693
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7110885381698608,
+      "learning_rate": 4.717578641289982e-05,
+      "loss": 1.2719,
+      "step": 10694
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5390009880065918,
+      "learning_rate": 4.717552207242929e-05,
+      "loss": 2.2441,
+      "step": 10695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.769260048866272,
+      "learning_rate": 4.717525773195877e-05,
+      "loss": 1.3313,
+      "step": 10696
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.747208595275879,
+      "learning_rate": 4.7174993391488236e-05,
+      "loss": 1.628,
+      "step": 10697
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7237541675567627,
+      "learning_rate": 4.7174729051017716e-05,
+      "loss": 1.3911,
+      "step": 10698
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8089679479599,
+      "learning_rate": 4.717446471054719e-05,
+      "loss": 1.3779,
+      "step": 10699
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9282413721084595,
+      "learning_rate": 4.717420037007666e-05,
+      "loss": 1.5264,
+      "step": 10700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1276497840881348,
+      "learning_rate": 4.717393602960614e-05,
+      "loss": 1.3997,
+      "step": 10701
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7361977100372314,
+      "learning_rate": 4.7173671689135604e-05,
+      "loss": 1.4989,
+      "step": 10702
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7247306108474731,
+      "learning_rate": 4.7173407348665084e-05,
+      "loss": 1.4017,
+      "step": 10703
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.438666820526123,
+      "learning_rate": 4.717314300819456e-05,
+      "loss": 1.517,
+      "step": 10704
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.903493046760559,
+      "learning_rate": 4.717287866772403e-05,
+      "loss": 1.371,
+      "step": 10705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0702545642852783,
+      "learning_rate": 4.7172614327253505e-05,
+      "loss": 1.5785,
+      "step": 10706
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5617420673370361,
+      "learning_rate": 4.717234998678298e-05,
+      "loss": 1.4398,
+      "step": 10707
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.715928554534912,
+      "learning_rate": 4.717208564631245e-05,
+      "loss": 1.547,
+      "step": 10708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.587226390838623,
+      "learning_rate": 4.717182130584193e-05,
+      "loss": 1.3703,
+      "step": 10709
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.828155517578125,
+      "learning_rate": 4.71715569653714e-05,
+      "loss": 1.3453,
+      "step": 10710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.185023546218872,
+      "learning_rate": 4.717129262490087e-05,
+      "loss": 2.4099,
+      "step": 10711
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6378709077835083,
+      "learning_rate": 4.7171028284430346e-05,
+      "loss": 1.3182,
+      "step": 10712
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7484115362167358,
+      "learning_rate": 4.717076394395982e-05,
+      "loss": 1.4758,
+      "step": 10713
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4143869876861572,
+      "learning_rate": 4.71704996034893e-05,
+      "loss": 1.5565,
+      "step": 10714
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.124466896057129,
+      "learning_rate": 4.717023526301877e-05,
+      "loss": 1.3064,
+      "step": 10715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.294473171234131,
+      "learning_rate": 4.716997092254825e-05,
+      "loss": 1.449,
+      "step": 10716
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5646591186523438,
+      "learning_rate": 4.7169706582077714e-05,
+      "loss": 1.4279,
+      "step": 10717
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9528683423995972,
+      "learning_rate": 4.7169442241607194e-05,
+      "loss": 1.4366,
+      "step": 10718
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.366757869720459,
+      "learning_rate": 4.716917790113667e-05,
+      "loss": 1.3395,
+      "step": 10719
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.492392659187317,
+      "learning_rate": 4.716891356066614e-05,
+      "loss": 1.3471,
+      "step": 10720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6593502759933472,
+      "learning_rate": 4.7168649220195615e-05,
+      "loss": 1.4817,
+      "step": 10721
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7597551345825195,
+      "learning_rate": 4.716838487972509e-05,
+      "loss": 1.4286,
+      "step": 10722
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.096071481704712,
+      "learning_rate": 4.716812053925456e-05,
+      "loss": 1.4347,
+      "step": 10723
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.765061855316162,
+      "learning_rate": 4.7167856198784036e-05,
+      "loss": 1.5263,
+      "step": 10724
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1457409858703613,
+      "learning_rate": 4.716759185831351e-05,
+      "loss": 1.5153,
+      "step": 10725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4425214529037476,
+      "learning_rate": 4.716732751784298e-05,
+      "loss": 1.3363,
+      "step": 10726
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.011403799057007,
+      "learning_rate": 4.7167063177372456e-05,
+      "loss": 1.5733,
+      "step": 10727
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.728007435798645,
+      "learning_rate": 4.716679883690193e-05,
+      "loss": 1.4434,
+      "step": 10728
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7717219591140747,
+      "learning_rate": 4.716653449643141e-05,
+      "loss": 1.3919,
+      "step": 10729
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6082454919815063,
+      "learning_rate": 4.716627015596088e-05,
+      "loss": 1.4249,
+      "step": 10730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5517914295196533,
+      "learning_rate": 4.716600581549036e-05,
+      "loss": 1.337,
+      "step": 10731
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15959095954895,
+      "learning_rate": 4.7165741475019824e-05,
+      "loss": 1.535,
+      "step": 10732
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7209352254867554,
+      "learning_rate": 4.7165477134549305e-05,
+      "loss": 1.3784,
+      "step": 10733
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4998995065689087,
+      "learning_rate": 4.716521279407878e-05,
+      "loss": 1.3325,
+      "step": 10734
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.490020513534546,
+      "learning_rate": 4.716494845360825e-05,
+      "loss": 1.2086,
+      "step": 10735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9366612434387207,
+      "learning_rate": 4.7164684113137725e-05,
+      "loss": 1.5492,
+      "step": 10736
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7727525234222412,
+      "learning_rate": 4.716441977266719e-05,
+      "loss": 1.3872,
+      "step": 10737
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1963117122650146,
+      "learning_rate": 4.716415543219667e-05,
+      "loss": 1.5305,
+      "step": 10738
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9017397165298462,
+      "learning_rate": 4.7163891091726146e-05,
+      "loss": 1.8353,
+      "step": 10739
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6892579793930054,
+      "learning_rate": 4.716362675125562e-05,
+      "loss": 1.256,
+      "step": 10740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.532572627067566,
+      "learning_rate": 4.716336241078509e-05,
+      "loss": 1.3411,
+      "step": 10741
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8850542306900024,
+      "learning_rate": 4.716309807031457e-05,
+      "loss": 1.4443,
+      "step": 10742
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.541368842124939,
+      "learning_rate": 4.716283372984404e-05,
+      "loss": 1.3751,
+      "step": 10743
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6348097324371338,
+      "learning_rate": 4.716256938937352e-05,
+      "loss": 1.3522,
+      "step": 10744
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3803353309631348,
+      "learning_rate": 4.716230504890299e-05,
+      "loss": 1.5238,
+      "step": 10745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6586462259292603,
+      "learning_rate": 4.716204070843246e-05,
+      "loss": 1.3829,
+      "step": 10746
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2281055450439453,
+      "learning_rate": 4.7161776367961935e-05,
+      "loss": 1.4828,
+      "step": 10747
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.417851448059082,
+      "learning_rate": 4.716151202749141e-05,
+      "loss": 1.2341,
+      "step": 10748
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4402358531951904,
+      "learning_rate": 4.716124768702089e-05,
+      "loss": 1.385,
+      "step": 10749
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7309670448303223,
+      "learning_rate": 4.7160983346550355e-05,
+      "loss": 1.2902,
+      "step": 10750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5219635963439941,
+      "learning_rate": 4.7160719006079836e-05,
+      "loss": 1.2918,
+      "step": 10751
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8200881481170654,
+      "learning_rate": 4.71604546656093e-05,
+      "loss": 1.347,
+      "step": 10752
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.853500485420227,
+      "learning_rate": 4.716019032513878e-05,
+      "loss": 1.398,
+      "step": 10753
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.475581645965576,
+      "learning_rate": 4.7159925984668256e-05,
+      "loss": 1.5805,
+      "step": 10754
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0515568256378174,
+      "learning_rate": 4.715966164419773e-05,
+      "loss": 1.5451,
+      "step": 10755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5962445735931396,
+      "learning_rate": 4.7159397303727204e-05,
+      "loss": 1.3137,
+      "step": 10756
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.660054087638855,
+      "learning_rate": 4.715913296325668e-05,
+      "loss": 1.7145,
+      "step": 10757
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5495933294296265,
+      "learning_rate": 4.715886862278615e-05,
+      "loss": 1.23,
+      "step": 10758
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.137254238128662,
+      "learning_rate": 4.7158604282315624e-05,
+      "loss": 1.6608,
+      "step": 10759
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7275437116622925,
+      "learning_rate": 4.71583399418451e-05,
+      "loss": 1.3714,
+      "step": 10760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.056586503982544,
+      "learning_rate": 4.715807560137457e-05,
+      "loss": 1.3414,
+      "step": 10761
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.070098876953125,
+      "learning_rate": 4.7157811260904045e-05,
+      "loss": 1.4896,
+      "step": 10762
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9184173345565796,
+      "learning_rate": 4.715754692043352e-05,
+      "loss": 1.4105,
+      "step": 10763
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.527970314025879,
+      "learning_rate": 4.7157282579963e-05,
+      "loss": 1.2812,
+      "step": 10764
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.578223705291748,
+      "learning_rate": 4.7157018239492466e-05,
+      "loss": 1.4776,
+      "step": 10765
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8737680912017822,
+      "learning_rate": 4.7156753899021946e-05,
+      "loss": 1.2728,
+      "step": 10766
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.095938205718994,
+      "learning_rate": 4.715648955855141e-05,
+      "loss": 1.5195,
+      "step": 10767
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0397067070007324,
+      "learning_rate": 4.715622521808089e-05,
+      "loss": 1.5758,
+      "step": 10768
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9184576272964478,
+      "learning_rate": 4.715596087761037e-05,
+      "loss": 1.3807,
+      "step": 10769
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.764762282371521,
+      "learning_rate": 4.715569653713984e-05,
+      "loss": 1.3037,
+      "step": 10770
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4603726863861084,
+      "learning_rate": 4.7155432196669314e-05,
+      "loss": 1.2942,
+      "step": 10771
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6574093103408813,
+      "learning_rate": 4.715516785619879e-05,
+      "loss": 1.3762,
+      "step": 10772
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9997081756591797,
+      "learning_rate": 4.715490351572826e-05,
+      "loss": 2.5509,
+      "step": 10773
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2487993240356445,
+      "learning_rate": 4.7154639175257735e-05,
+      "loss": 1.4998,
+      "step": 10774
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.229078531265259,
+      "learning_rate": 4.715437483478721e-05,
+      "loss": 1.2878,
+      "step": 10775
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.279452323913574,
+      "learning_rate": 4.715411049431668e-05,
+      "loss": 1.5939,
+      "step": 10776
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7148311138153076,
+      "learning_rate": 4.7153846153846155e-05,
+      "loss": 1.4095,
+      "step": 10777
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.016239881515503,
+      "learning_rate": 4.715358181337563e-05,
+      "loss": 1.4672,
+      "step": 10778
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7139012813568115,
+      "learning_rate": 4.715331747290511e-05,
+      "loss": 1.3053,
+      "step": 10779
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.737443208694458,
+      "learning_rate": 4.7153053132434576e-05,
+      "loss": 1.5838,
+      "step": 10780
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.218427896499634,
+      "learning_rate": 4.7152788791964056e-05,
+      "loss": 1.3618,
+      "step": 10781
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5295640230178833,
+      "learning_rate": 4.715252445149352e-05,
+      "loss": 1.3539,
+      "step": 10782
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5247915983200073,
+      "learning_rate": 4.7152260111023e-05,
+      "loss": 1.2782,
+      "step": 10783
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7564772367477417,
+      "learning_rate": 4.715199577055248e-05,
+      "loss": 1.4296,
+      "step": 10784
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7046327590942383,
+      "learning_rate": 4.7151731430081944e-05,
+      "loss": 2.1986,
+      "step": 10785
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4823377132415771,
+      "learning_rate": 4.7151467089611424e-05,
+      "loss": 1.7446,
+      "step": 10786
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4188029766082764,
+      "learning_rate": 4.715120274914089e-05,
+      "loss": 1.4929,
+      "step": 10787
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.657086730003357,
+      "learning_rate": 4.715093840867037e-05,
+      "loss": 1.4294,
+      "step": 10788
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6690855026245117,
+      "learning_rate": 4.7150674068199845e-05,
+      "loss": 1.3416,
+      "step": 10789
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1234805583953857,
+      "learning_rate": 4.715040972772932e-05,
+      "loss": 1.4606,
+      "step": 10790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2912856340408325,
+      "learning_rate": 4.715014538725879e-05,
+      "loss": 1.252,
+      "step": 10791
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8252959251403809,
+      "learning_rate": 4.7149881046788266e-05,
+      "loss": 1.29,
+      "step": 10792
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7290743589401245,
+      "learning_rate": 4.714961670631774e-05,
+      "loss": 1.4789,
+      "step": 10793
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.924736738204956,
+      "learning_rate": 4.714935236584721e-05,
+      "loss": 1.5359,
+      "step": 10794
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6417828798294067,
+      "learning_rate": 4.7149088025376686e-05,
+      "loss": 1.367,
+      "step": 10795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.889129638671875,
+      "learning_rate": 4.714882368490616e-05,
+      "loss": 1.5295,
+      "step": 10796
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.544944167137146,
+      "learning_rate": 4.7148559344435634e-05,
+      "loss": 1.3901,
+      "step": 10797
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8914906978607178,
+      "learning_rate": 4.714829500396511e-05,
+      "loss": 1.481,
+      "step": 10798
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1542131900787354,
+      "learning_rate": 4.714803066349459e-05,
+      "loss": 1.4443,
+      "step": 10799
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.430532932281494,
+      "learning_rate": 4.7147766323024054e-05,
+      "loss": 1.4929,
+      "step": 10800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8928592205047607,
+      "learning_rate": 4.7147501982553535e-05,
+      "loss": 1.3976,
+      "step": 10801
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.500723958015442,
+      "learning_rate": 4.7147237642083e-05,
+      "loss": 1.2438,
+      "step": 10802
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4343994855880737,
+      "learning_rate": 4.714697330161248e-05,
+      "loss": 1.83,
+      "step": 10803
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.646747350692749,
+      "learning_rate": 4.7146708961141955e-05,
+      "loss": 1.5492,
+      "step": 10804
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9493449926376343,
+      "learning_rate": 4.714644462067143e-05,
+      "loss": 1.4904,
+      "step": 10805
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2268319129943848,
+      "learning_rate": 4.71461802802009e-05,
+      "loss": 1.3584,
+      "step": 10806
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.16373610496521,
+      "learning_rate": 4.7145915939730376e-05,
+      "loss": 1.5196,
+      "step": 10807
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5577313899993896,
+      "learning_rate": 4.714565159925985e-05,
+      "loss": 1.4399,
+      "step": 10808
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7113337516784668,
+      "learning_rate": 4.714538725878932e-05,
+      "loss": 1.4776,
+      "step": 10809
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.044391632080078,
+      "learning_rate": 4.71451229183188e-05,
+      "loss": 1.4382,
+      "step": 10810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.61420476436615,
+      "learning_rate": 4.714485857784827e-05,
+      "loss": 1.3285,
+      "step": 10811
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0227091312408447,
+      "learning_rate": 4.7144594237377744e-05,
+      "loss": 1.4281,
+      "step": 10812
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5869730710983276,
+      "learning_rate": 4.714432989690722e-05,
+      "loss": 1.2919,
+      "step": 10813
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0438013076782227,
+      "learning_rate": 4.71440655564367e-05,
+      "loss": 1.419,
+      "step": 10814
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5807368755340576,
+      "learning_rate": 4.7143801215966165e-05,
+      "loss": 1.3857,
+      "step": 10815
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.931030035018921,
+      "learning_rate": 4.7143536875495645e-05,
+      "loss": 1.3744,
+      "step": 10816
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.243324041366577,
+      "learning_rate": 4.714327253502511e-05,
+      "loss": 1.4124,
+      "step": 10817
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7330104112625122,
+      "learning_rate": 4.714300819455459e-05,
+      "loss": 1.357,
+      "step": 10818
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7962526082992554,
+      "learning_rate": 4.7142743854084066e-05,
+      "loss": 1.4065,
+      "step": 10819
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.069589138031006,
+      "learning_rate": 4.714247951361353e-05,
+      "loss": 1.5138,
+      "step": 10820
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.646064043045044,
+      "learning_rate": 4.714221517314301e-05,
+      "loss": 1.2696,
+      "step": 10821
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453789710998535,
+      "learning_rate": 4.714195083267248e-05,
+      "loss": 1.4505,
+      "step": 10822
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8233917951583862,
+      "learning_rate": 4.714168649220196e-05,
+      "loss": 1.4356,
+      "step": 10823
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6206943988800049,
+      "learning_rate": 4.7141422151731434e-05,
+      "loss": 1.2713,
+      "step": 10824
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8745158910751343,
+      "learning_rate": 4.714115781126091e-05,
+      "loss": 1.3688,
+      "step": 10825
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7955198287963867,
+      "learning_rate": 4.714089347079038e-05,
+      "loss": 1.4522,
+      "step": 10826
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8169372081756592,
+      "learning_rate": 4.7140629130319854e-05,
+      "loss": 1.3767,
+      "step": 10827
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.867360234260559,
+      "learning_rate": 4.714036478984933e-05,
+      "loss": 1.5478,
+      "step": 10828
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8371185064315796,
+      "learning_rate": 4.71401004493788e-05,
+      "loss": 1.502,
+      "step": 10829
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4835273027420044,
+      "learning_rate": 4.7139836108908275e-05,
+      "loss": 1.3019,
+      "step": 10830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6123006343841553,
+      "learning_rate": 4.713957176843775e-05,
+      "loss": 1.5138,
+      "step": 10831
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1064460277557373,
+      "learning_rate": 4.713930742796722e-05,
+      "loss": 1.6004,
+      "step": 10832
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.865803837776184,
+      "learning_rate": 4.7139043087496696e-05,
+      "loss": 1.4829,
+      "step": 10833
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8277639150619507,
+      "learning_rate": 4.7138778747026176e-05,
+      "loss": 1.3025,
+      "step": 10834
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.310964345932007,
+      "learning_rate": 4.713851440655564e-05,
+      "loss": 1.2331,
+      "step": 10835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7228060960769653,
+      "learning_rate": 4.713825006608512e-05,
+      "loss": 1.4219,
+      "step": 10836
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8647209405899048,
+      "learning_rate": 4.713798572561459e-05,
+      "loss": 2.0574,
+      "step": 10837
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7803635597229004,
+      "learning_rate": 4.713772138514407e-05,
+      "loss": 1.3552,
+      "step": 10838
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4320932626724243,
+      "learning_rate": 4.7137457044673544e-05,
+      "loss": 1.2214,
+      "step": 10839
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4157652854919434,
+      "learning_rate": 4.713719270420302e-05,
+      "loss": 1.5447,
+      "step": 10840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7358165979385376,
+      "learning_rate": 4.713692836373249e-05,
+      "loss": 1.361,
+      "step": 10841
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9332820177078247,
+      "learning_rate": 4.7136664023261965e-05,
+      "loss": 1.3681,
+      "step": 10842
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7034341096878052,
+      "learning_rate": 4.713639968279144e-05,
+      "loss": 1.5058,
+      "step": 10843
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8128514289855957,
+      "learning_rate": 4.713613534232091e-05,
+      "loss": 1.2741,
+      "step": 10844
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6915851831436157,
+      "learning_rate": 4.7135871001850385e-05,
+      "loss": 1.3584,
+      "step": 10845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.925041913986206,
+      "learning_rate": 4.713560666137986e-05,
+      "loss": 1.4172,
+      "step": 10846
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7380731105804443,
+      "learning_rate": 4.713534232090933e-05,
+      "loss": 1.3838,
+      "step": 10847
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.951303482055664,
+      "learning_rate": 4.7135077980438806e-05,
+      "loss": 1.4971,
+      "step": 10848
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6136267185211182,
+      "learning_rate": 4.7134813639968286e-05,
+      "loss": 1.2644,
+      "step": 10849
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9299705028533936,
+      "learning_rate": 4.713454929949775e-05,
+      "loss": 1.4769,
+      "step": 10850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0092835426330566,
+      "learning_rate": 4.7134284959027233e-05,
+      "loss": 1.5191,
+      "step": 10851
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7385051250457764,
+      "learning_rate": 4.71340206185567e-05,
+      "loss": 1.2541,
+      "step": 10852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.148106813430786,
+      "learning_rate": 4.713375627808618e-05,
+      "loss": 1.3179,
+      "step": 10853
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8637431859970093,
+      "learning_rate": 4.7133491937615654e-05,
+      "loss": 1.303,
+      "step": 10854
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7904574871063232,
+      "learning_rate": 4.713322759714512e-05,
+      "loss": 1.5093,
+      "step": 10855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8819538354873657,
+      "learning_rate": 4.71329632566746e-05,
+      "loss": 1.5061,
+      "step": 10856
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.155397653579712,
+      "learning_rate": 4.713269891620407e-05,
+      "loss": 1.6464,
+      "step": 10857
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6858412027359009,
+      "learning_rate": 4.713243457573355e-05,
+      "loss": 1.4708,
+      "step": 10858
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.52125883102417,
+      "learning_rate": 4.713217023526302e-05,
+      "loss": 1.3185,
+      "step": 10859
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0990211963653564,
+      "learning_rate": 4.7131905894792496e-05,
+      "loss": 2.1834,
+      "step": 10860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.79714834690094,
+      "learning_rate": 4.713164155432197e-05,
+      "loss": 1.5818,
+      "step": 10861
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6920104026794434,
+      "learning_rate": 4.713137721385144e-05,
+      "loss": 1.3228,
+      "step": 10862
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3622502088546753,
+      "learning_rate": 4.7131112873380916e-05,
+      "loss": 1.108,
+      "step": 10863
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8092656135559082,
+      "learning_rate": 4.713084853291039e-05,
+      "loss": 1.4027,
+      "step": 10864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7396240234375,
+      "learning_rate": 4.7130584192439863e-05,
+      "loss": 1.4708,
+      "step": 10865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0056467056274414,
+      "learning_rate": 4.713031985196934e-05,
+      "loss": 1.522,
+      "step": 10866
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8539036512374878,
+      "learning_rate": 4.713005551149881e-05,
+      "loss": 1.2967,
+      "step": 10867
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4107807874679565,
+      "learning_rate": 4.7129791171028284e-05,
+      "loss": 1.7893,
+      "step": 10868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9281874895095825,
+      "learning_rate": 4.7129526830557765e-05,
+      "loss": 1.5172,
+      "step": 10869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5810186862945557,
+      "learning_rate": 4.712926249008723e-05,
+      "loss": 1.3236,
+      "step": 10870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4388374090194702,
+      "learning_rate": 4.712899814961671e-05,
+      "loss": 1.2484,
+      "step": 10871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8878505229949951,
+      "learning_rate": 4.712873380914618e-05,
+      "loss": 2.5086,
+      "step": 10872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015432596206665,
+      "learning_rate": 4.712846946867566e-05,
+      "loss": 1.3729,
+      "step": 10873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8390600681304932,
+      "learning_rate": 4.712820512820513e-05,
+      "loss": 1.3493,
+      "step": 10874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4471842050552368,
+      "learning_rate": 4.7127940787734606e-05,
+      "loss": 1.227,
+      "step": 10875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6256604194641113,
+      "learning_rate": 4.712767644726408e-05,
+      "loss": 1.4864,
+      "step": 10876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4431490898132324,
+      "learning_rate": 4.712741210679355e-05,
+      "loss": 1.3303,
+      "step": 10877
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7829314470291138,
+      "learning_rate": 4.712714776632303e-05,
+      "loss": 1.4752,
+      "step": 10878
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7366926670074463,
+      "learning_rate": 4.71268834258525e-05,
+      "loss": 2.3,
+      "step": 10879
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8693104982376099,
+      "learning_rate": 4.7126619085381974e-05,
+      "loss": 1.3346,
+      "step": 10880
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.188426971435547,
+      "learning_rate": 4.712635474491145e-05,
+      "loss": 1.5611,
+      "step": 10881
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6044211387634277,
+      "learning_rate": 4.712609040444092e-05,
+      "loss": 1.3302,
+      "step": 10882
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7411069869995117,
+      "learning_rate": 4.7125826063970395e-05,
+      "loss": 1.5004,
+      "step": 10883
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6897813081741333,
+      "learning_rate": 4.7125561723499875e-05,
+      "loss": 2.1386,
+      "step": 10884
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.566445231437683,
+      "learning_rate": 4.712529738302934e-05,
+      "loss": 1.488,
+      "step": 10885
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4274728298187256,
+      "learning_rate": 4.712503304255882e-05,
+      "loss": 1.5573,
+      "step": 10886
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0503311157226562,
+      "learning_rate": 4.712476870208829e-05,
+      "loss": 2.4582,
+      "step": 10887
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.160536289215088,
+      "learning_rate": 4.712450436161777e-05,
+      "loss": 1.3357,
+      "step": 10888
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8825273513793945,
+      "learning_rate": 4.712424002114724e-05,
+      "loss": 1.545,
+      "step": 10889
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.418781280517578,
+      "learning_rate": 4.7123975680676716e-05,
+      "loss": 1.5245,
+      "step": 10890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8557369709014893,
+      "learning_rate": 4.712371134020619e-05,
+      "loss": 1.4686,
+      "step": 10891
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.345921277999878,
+      "learning_rate": 4.712344699973566e-05,
+      "loss": 1.4967,
+      "step": 10892
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4920192956924438,
+      "learning_rate": 4.712318265926514e-05,
+      "loss": 1.3573,
+      "step": 10893
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5648586750030518,
+      "learning_rate": 4.712291831879461e-05,
+      "loss": 1.3564,
+      "step": 10894
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9039826393127441,
+      "learning_rate": 4.7122653978324084e-05,
+      "loss": 1.5284,
+      "step": 10895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1551945209503174,
+      "learning_rate": 4.712238963785356e-05,
+      "loss": 1.4295,
+      "step": 10896
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5359561443328857,
+      "learning_rate": 4.712212529738303e-05,
+      "loss": 1.4671,
+      "step": 10897
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.303654432296753,
+      "learning_rate": 4.7121860956912505e-05,
+      "loss": 1.3833,
+      "step": 10898
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9739184379577637,
+      "learning_rate": 4.7121596616441985e-05,
+      "loss": 1.7227,
+      "step": 10899
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2236316204071045,
+      "learning_rate": 4.712133227597145e-05,
+      "loss": 1.423,
+      "step": 10900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7104859352111816,
+      "learning_rate": 4.7121067935500926e-05,
+      "loss": 1.3825,
+      "step": 10901
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9077966213226318,
+      "learning_rate": 4.71208035950304e-05,
+      "loss": 1.4308,
+      "step": 10902
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.306807041168213,
+      "learning_rate": 4.712053925455987e-05,
+      "loss": 1.4795,
+      "step": 10903
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7755224704742432,
+      "learning_rate": 4.712027491408935e-05,
+      "loss": 1.4566,
+      "step": 10904
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5616408586502075,
+      "learning_rate": 4.712001057361882e-05,
+      "loss": 1.2421,
+      "step": 10905
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5108366012573242,
+      "learning_rate": 4.71197462331483e-05,
+      "loss": 1.2471,
+      "step": 10906
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4387794733047485,
+      "learning_rate": 4.711948189267777e-05,
+      "loss": 1.3457,
+      "step": 10907
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.558166742324829,
+      "learning_rate": 4.711921755220725e-05,
+      "loss": 1.0433,
+      "step": 10908
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9438709020614624,
+      "learning_rate": 4.711895321173672e-05,
+      "loss": 1.2321,
+      "step": 10909
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9372243881225586,
+      "learning_rate": 4.7118688871266194e-05,
+      "loss": 1.3909,
+      "step": 10910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8646689653396606,
+      "learning_rate": 4.711842453079567e-05,
+      "loss": 1.2899,
+      "step": 10911
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6674591302871704,
+      "learning_rate": 4.711816019032514e-05,
+      "loss": 1.4137,
+      "step": 10912
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6903283596038818,
+      "learning_rate": 4.7117895849854615e-05,
+      "loss": 1.4999,
+      "step": 10913
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6724259853363037,
+      "learning_rate": 4.711763150938409e-05,
+      "loss": 1.4903,
+      "step": 10914
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5955485105514526,
+      "learning_rate": 4.711736716891356e-05,
+      "loss": 1.3103,
+      "step": 10915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9250894784927368,
+      "learning_rate": 4.7117102828443036e-05,
+      "loss": 1.4444,
+      "step": 10916
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2162184715270996,
+      "learning_rate": 4.711683848797251e-05,
+      "loss": 1.6066,
+      "step": 10917
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.531352996826172,
+      "learning_rate": 4.711657414750198e-05,
+      "loss": 1.517,
+      "step": 10918
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0601916313171387,
+      "learning_rate": 4.7116309807031463e-05,
+      "loss": 1.421,
+      "step": 10919
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2695624828338623,
+      "learning_rate": 4.711604546656093e-05,
+      "loss": 1.3909,
+      "step": 10920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6747759580612183,
+      "learning_rate": 4.711578112609041e-05,
+      "loss": 1.2673,
+      "step": 10921
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0903263092041016,
+      "learning_rate": 4.711551678561988e-05,
+      "loss": 1.4352,
+      "step": 10922
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.571589946746826,
+      "learning_rate": 4.711525244514936e-05,
+      "loss": 1.6587,
+      "step": 10923
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.453572988510132,
+      "learning_rate": 4.711498810467883e-05,
+      "loss": 1.4364,
+      "step": 10924
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.554404854774475,
+      "learning_rate": 4.7114723764208305e-05,
+      "loss": 1.4112,
+      "step": 10925
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7884461879730225,
+      "learning_rate": 4.711445942373778e-05,
+      "loss": 2.6798,
+      "step": 10926
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6688662767410278,
+      "learning_rate": 4.7114195083267245e-05,
+      "loss": 1.3825,
+      "step": 10927
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9372128248214722,
+      "learning_rate": 4.7113930742796726e-05,
+      "loss": 1.4197,
+      "step": 10928
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.114471673965454,
+      "learning_rate": 4.71136664023262e-05,
+      "loss": 1.3171,
+      "step": 10929
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6251312494277954,
+      "learning_rate": 4.711340206185567e-05,
+      "loss": 1.4246,
+      "step": 10930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8771642446517944,
+      "learning_rate": 4.7113137721385146e-05,
+      "loss": 1.3199,
+      "step": 10931
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7563910484313965,
+      "learning_rate": 4.711287338091462e-05,
+      "loss": 1.2799,
+      "step": 10932
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6461074352264404,
+      "learning_rate": 4.7112609040444093e-05,
+      "loss": 1.2771,
+      "step": 10933
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6156420707702637,
+      "learning_rate": 4.7112344699973574e-05,
+      "loss": 1.2892,
+      "step": 10934
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.131526470184326,
+      "learning_rate": 4.711208035950304e-05,
+      "loss": 1.3865,
+      "step": 10935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8629403114318848,
+      "learning_rate": 4.711181601903252e-05,
+      "loss": 1.3555,
+      "step": 10936
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.181396245956421,
+      "learning_rate": 4.711155167856199e-05,
+      "loss": 1.4846,
+      "step": 10937
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7636953592300415,
+      "learning_rate": 4.711128733809146e-05,
+      "loss": 1.5081,
+      "step": 10938
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7144885063171387,
+      "learning_rate": 4.711102299762094e-05,
+      "loss": 1.4185,
+      "step": 10939
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7934317588806152,
+      "learning_rate": 4.711075865715041e-05,
+      "loss": 1.3391,
+      "step": 10940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7369014024734497,
+      "learning_rate": 4.711049431667989e-05,
+      "loss": 1.4726,
+      "step": 10941
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8259104490280151,
+      "learning_rate": 4.7110229976209356e-05,
+      "loss": 1.4682,
+      "step": 10942
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5380771160125732,
+      "learning_rate": 4.7109965635738836e-05,
+      "loss": 1.3624,
+      "step": 10943
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.815024495124817,
+      "learning_rate": 4.710970129526831e-05,
+      "loss": 1.3857,
+      "step": 10944
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.130171298980713,
+      "learning_rate": 4.710943695479778e-05,
+      "loss": 1.1563,
+      "step": 10945
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.691319227218628,
+      "learning_rate": 4.7109172614327257e-05,
+      "loss": 1.2142,
+      "step": 10946
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7300230264663696,
+      "learning_rate": 4.710890827385673e-05,
+      "loss": 1.3566,
+      "step": 10947
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8762669563293457,
+      "learning_rate": 4.7108643933386204e-05,
+      "loss": 1.521,
+      "step": 10948
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.97959566116333,
+      "learning_rate": 4.710837959291568e-05,
+      "loss": 1.487,
+      "step": 10949
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6974661350250244,
+      "learning_rate": 4.710811525244515e-05,
+      "loss": 1.3251,
+      "step": 10950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.888310432434082,
+      "learning_rate": 4.7107850911974624e-05,
+      "loss": 1.4091,
+      "step": 10951
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.969069004058838,
+      "learning_rate": 4.71075865715041e-05,
+      "loss": 1.487,
+      "step": 10952
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6105111837387085,
+      "learning_rate": 4.710732223103357e-05,
+      "loss": 1.413,
+      "step": 10953
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6157790422439575,
+      "learning_rate": 4.710705789056305e-05,
+      "loss": 1.2392,
+      "step": 10954
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6065138578414917,
+      "learning_rate": 4.710679355009252e-05,
+      "loss": 1.3964,
+      "step": 10955
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.748921275138855,
+      "learning_rate": 4.7106529209622e-05,
+      "loss": 1.3737,
+      "step": 10956
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.736341953277588,
+      "learning_rate": 4.7106264869151466e-05,
+      "loss": 1.4888,
+      "step": 10957
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6314929723739624,
+      "learning_rate": 4.7106000528680946e-05,
+      "loss": 1.2576,
+      "step": 10958
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6557190418243408,
+      "learning_rate": 4.710573618821042e-05,
+      "loss": 1.3052,
+      "step": 10959
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2573652267456055,
+      "learning_rate": 4.710547184773989e-05,
+      "loss": 1.5584,
+      "step": 10960
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3905984163284302,
+      "learning_rate": 4.710520750726937e-05,
+      "loss": 1.2613,
+      "step": 10961
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3530938625335693,
+      "learning_rate": 4.710494316679884e-05,
+      "loss": 1.3721,
+      "step": 10962
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2668869495391846,
+      "learning_rate": 4.7104678826328314e-05,
+      "loss": 1.2765,
+      "step": 10963
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9072965383529663,
+      "learning_rate": 4.710441448585779e-05,
+      "loss": 1.2365,
+      "step": 10964
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1667306423187256,
+      "learning_rate": 4.710415014538726e-05,
+      "loss": 1.7724,
+      "step": 10965
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5872372388839722,
+      "learning_rate": 4.7103885804916735e-05,
+      "loss": 1.2731,
+      "step": 10966
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9180704355239868,
+      "learning_rate": 4.710362146444621e-05,
+      "loss": 1.5887,
+      "step": 10967
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0299508571624756,
+      "learning_rate": 4.710335712397568e-05,
+      "loss": 1.4512,
+      "step": 10968
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3257932662963867,
+      "learning_rate": 4.710309278350516e-05,
+      "loss": 1.1756,
+      "step": 10969
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8244216442108154,
+      "learning_rate": 4.710282844303463e-05,
+      "loss": 1.5201,
+      "step": 10970
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8688292503356934,
+      "learning_rate": 4.710256410256411e-05,
+      "loss": 1.4161,
+      "step": 10971
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3840484619140625,
+      "learning_rate": 4.7102299762093576e-05,
+      "loss": 1.6779,
+      "step": 10972
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.854941725730896,
+      "learning_rate": 4.710203542162305e-05,
+      "loss": 1.4771,
+      "step": 10973
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8714083433151245,
+      "learning_rate": 4.710177108115253e-05,
+      "loss": 1.6183,
+      "step": 10974
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8930537700653076,
+      "learning_rate": 4.7101506740682e-05,
+      "loss": 1.6838,
+      "step": 10975
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6154173612594604,
+      "learning_rate": 4.710124240021148e-05,
+      "loss": 1.3274,
+      "step": 10976
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.546767234802246,
+      "learning_rate": 4.7100978059740944e-05,
+      "loss": 1.2992,
+      "step": 10977
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5518637895584106,
+      "learning_rate": 4.7100713719270424e-05,
+      "loss": 1.3092,
+      "step": 10978
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.889849305152893,
+      "learning_rate": 4.71004493787999e-05,
+      "loss": 1.4812,
+      "step": 10979
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.75383722782135,
+      "learning_rate": 4.710018503832937e-05,
+      "loss": 1.4337,
+      "step": 10980
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7045044898986816,
+      "learning_rate": 4.7099920697858845e-05,
+      "loss": 1.4943,
+      "step": 10981
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0436885356903076,
+      "learning_rate": 4.709965635738832e-05,
+      "loss": 1.4521,
+      "step": 10982
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7896027565002441,
+      "learning_rate": 4.709939201691779e-05,
+      "loss": 1.3951,
+      "step": 10983
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7125235795974731,
+      "learning_rate": 4.7099127676447266e-05,
+      "loss": 1.2542,
+      "step": 10984
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4092600345611572,
+      "learning_rate": 4.709886333597674e-05,
+      "loss": 1.1936,
+      "step": 10985
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6442378759384155,
+      "learning_rate": 4.709859899550621e-05,
+      "loss": 1.4864,
+      "step": 10986
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.494064450263977,
+      "learning_rate": 4.7098334655035687e-05,
+      "loss": 1.2772,
+      "step": 10987
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.130387544631958,
+      "learning_rate": 4.709807031456516e-05,
+      "loss": 1.4823,
+      "step": 10988
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8758803606033325,
+      "learning_rate": 4.709780597409464e-05,
+      "loss": 1.4154,
+      "step": 10989
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.68186616897583,
+      "learning_rate": 4.709754163362411e-05,
+      "loss": 1.5596,
+      "step": 10990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.346055269241333,
+      "learning_rate": 4.709727729315359e-05,
+      "loss": 1.4698,
+      "step": 10991
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3711657524108887,
+      "learning_rate": 4.7097012952683054e-05,
+      "loss": 1.3954,
+      "step": 10992
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5343767404556274,
+      "learning_rate": 4.7096748612212535e-05,
+      "loss": 1.3074,
+      "step": 10993
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1866848468780518,
+      "learning_rate": 4.709648427174201e-05,
+      "loss": 1.5389,
+      "step": 10994
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6191970109939575,
+      "learning_rate": 4.709621993127148e-05,
+      "loss": 1.2675,
+      "step": 10995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8140848875045776,
+      "learning_rate": 4.7095955590800955e-05,
+      "loss": 1.376,
+      "step": 10996
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.992633581161499,
+      "learning_rate": 4.709569125033043e-05,
+      "loss": 1.4722,
+      "step": 10997
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0445597171783447,
+      "learning_rate": 4.70954269098599e-05,
+      "loss": 1.3294,
+      "step": 10998
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2112412452697754,
+      "learning_rate": 4.7095162569389376e-05,
+      "loss": 1.3901,
+      "step": 10999
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109147071838379,
+      "learning_rate": 4.709489822891885e-05,
+      "loss": 1.4347,
+      "step": 11000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.520932197570801,
+      "learning_rate": 4.709463388844832e-05,
+      "loss": 1.4432,
+      "step": 11001
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7465118169784546,
+      "learning_rate": 4.70943695479778e-05,
+      "loss": 1.3497,
+      "step": 11002
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.641168236732483,
+      "learning_rate": 4.709410520750727e-05,
+      "loss": 1.3992,
+      "step": 11003
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7900891304016113,
+      "learning_rate": 4.709384086703675e-05,
+      "loss": 1.3873,
+      "step": 11004
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.587862968444824,
+      "learning_rate": 4.709357652656622e-05,
+      "loss": 1.4612,
+      "step": 11005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3924553394317627,
+      "learning_rate": 4.70933121860957e-05,
+      "loss": 1.3623,
+      "step": 11006
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7901774644851685,
+      "learning_rate": 4.7093047845625165e-05,
+      "loss": 1.4123,
+      "step": 11007
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4161529541015625,
+      "learning_rate": 4.7092783505154645e-05,
+      "loss": 1.2425,
+      "step": 11008
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0682551860809326,
+      "learning_rate": 4.709251916468412e-05,
+      "loss": 1.3966,
+      "step": 11009
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6172592639923096,
+      "learning_rate": 4.7092254824213585e-05,
+      "loss": 1.2295,
+      "step": 11010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.851180911064148,
+      "learning_rate": 4.7091990483743066e-05,
+      "loss": 1.473,
+      "step": 11011
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9731943607330322,
+      "learning_rate": 4.709172614327253e-05,
+      "loss": 1.3443,
+      "step": 11012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4670369625091553,
+      "learning_rate": 4.709146180280201e-05,
+      "loss": 1.38,
+      "step": 11013
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9821257591247559,
+      "learning_rate": 4.7091197462331487e-05,
+      "loss": 1.4846,
+      "step": 11014
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9254488945007324,
+      "learning_rate": 4.709093312186096e-05,
+      "loss": 1.42,
+      "step": 11015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0786962509155273,
+      "learning_rate": 4.7090668781390434e-05,
+      "loss": 1.4101,
+      "step": 11016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.778028130531311,
+      "learning_rate": 4.709040444091991e-05,
+      "loss": 1.4187,
+      "step": 11017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.715399146080017,
+      "learning_rate": 4.709014010044938e-05,
+      "loss": 1.364,
+      "step": 11018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.466064453125,
+      "learning_rate": 4.7089875759978854e-05,
+      "loss": 1.317,
+      "step": 11019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2032270431518555,
+      "learning_rate": 4.708961141950833e-05,
+      "loss": 1.3642,
+      "step": 11020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0007050037384033,
+      "learning_rate": 4.70893470790378e-05,
+      "loss": 1.4096,
+      "step": 11021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.466971755027771,
+      "learning_rate": 4.7089082738567275e-05,
+      "loss": 1.283,
+      "step": 11022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.229886531829834,
+      "learning_rate": 4.708881839809675e-05,
+      "loss": 2.7251,
+      "step": 11023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5575265884399414,
+      "learning_rate": 4.708855405762623e-05,
+      "loss": 1.3568,
+      "step": 11024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4654958248138428,
+      "learning_rate": 4.7088289717155696e-05,
+      "loss": 1.4708,
+      "step": 11025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8500728607177734,
+      "learning_rate": 4.7088025376685176e-05,
+      "loss": 1.377,
+      "step": 11026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.593885898590088,
+      "learning_rate": 4.708776103621464e-05,
+      "loss": 1.5507,
+      "step": 11027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9410592317581177,
+      "learning_rate": 4.708749669574412e-05,
+      "loss": 1.3134,
+      "step": 11028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4700385332107544,
+      "learning_rate": 4.70872323552736e-05,
+      "loss": 1.4213,
+      "step": 11029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.778092384338379,
+      "learning_rate": 4.708696801480307e-05,
+      "loss": 1.2498,
+      "step": 11030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8196563720703125,
+      "learning_rate": 4.7086703674332544e-05,
+      "loss": 1.4117,
+      "step": 11031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.604198932647705,
+      "learning_rate": 4.708643933386202e-05,
+      "loss": 1.3507,
+      "step": 11032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6089240312576294,
+      "learning_rate": 4.708617499339149e-05,
+      "loss": 1.4432,
+      "step": 11033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4915255308151245,
+      "learning_rate": 4.7085910652920965e-05,
+      "loss": 1.2701,
+      "step": 11034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7505300045013428,
+      "learning_rate": 4.708564631245044e-05,
+      "loss": 1.4088,
+      "step": 11035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.201558828353882,
+      "learning_rate": 4.708538197197991e-05,
+      "loss": 1.466,
+      "step": 11036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.789561152458191,
+      "learning_rate": 4.7085117631509385e-05,
+      "loss": 1.4184,
+      "step": 11037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6190178394317627,
+      "learning_rate": 4.708485329103886e-05,
+      "loss": 1.5024,
+      "step": 11038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5592067241668701,
+      "learning_rate": 4.708458895056834e-05,
+      "loss": 1.2833,
+      "step": 11039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8699793815612793,
+      "learning_rate": 4.7084324610097806e-05,
+      "loss": 1.4908,
+      "step": 11040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3983324766159058,
+      "learning_rate": 4.7084060269627286e-05,
+      "loss": 2.2268,
+      "step": 11041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2143168449401855,
+      "learning_rate": 4.708379592915675e-05,
+      "loss": 1.4993,
+      "step": 11042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9475653171539307,
+      "learning_rate": 4.7083531588686234e-05,
+      "loss": 1.3945,
+      "step": 11043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6735897064208984,
+      "learning_rate": 4.708326724821571e-05,
+      "loss": 1.5097,
+      "step": 11044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7080440521240234,
+      "learning_rate": 4.7083002907745174e-05,
+      "loss": 1.3587,
+      "step": 11045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6892242431640625,
+      "learning_rate": 4.7082738567274654e-05,
+      "loss": 1.096,
+      "step": 11046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9986776113510132,
+      "learning_rate": 4.708247422680412e-05,
+      "loss": 1.4236,
+      "step": 11047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6798917055130005,
+      "learning_rate": 4.70822098863336e-05,
+      "loss": 1.2748,
+      "step": 11048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5703645944595337,
+      "learning_rate": 4.7081945545863075e-05,
+      "loss": 1.4171,
+      "step": 11049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.634251594543457,
+      "learning_rate": 4.708168120539255e-05,
+      "loss": 1.2884,
+      "step": 11050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5060844421386719,
+      "learning_rate": 4.708141686492202e-05,
+      "loss": 1.3752,
+      "step": 11051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9179317951202393,
+      "learning_rate": 4.7081152524451496e-05,
+      "loss": 1.4529,
+      "step": 11052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.460073232650757,
+      "learning_rate": 4.708088818398097e-05,
+      "loss": 1.4926,
+      "step": 11053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.561052680015564,
+      "learning_rate": 4.708062384351045e-05,
+      "loss": 2.403,
+      "step": 11054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9428346157073975,
+      "learning_rate": 4.7080359503039916e-05,
+      "loss": 1.5051,
+      "step": 11055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4206230640411377,
+      "learning_rate": 4.708009516256939e-05,
+      "loss": 1.3493,
+      "step": 11056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7617220878601074,
+      "learning_rate": 4.7079830822098864e-05,
+      "loss": 1.3827,
+      "step": 11057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6153366565704346,
+      "learning_rate": 4.707956648162834e-05,
+      "loss": 1.4219,
+      "step": 11058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9331754446029663,
+      "learning_rate": 4.707930214115782e-05,
+      "loss": 1.2923,
+      "step": 11059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3677597045898438,
+      "learning_rate": 4.7079037800687284e-05,
+      "loss": 1.4485,
+      "step": 11060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3076858520507812,
+      "learning_rate": 4.7078773460216765e-05,
+      "loss": 1.5851,
+      "step": 11061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5548579692840576,
+      "learning_rate": 4.707850911974623e-05,
+      "loss": 1.9904,
+      "step": 11062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6462457180023193,
+      "learning_rate": 4.707824477927571e-05,
+      "loss": 1.4264,
+      "step": 11063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5120177268981934,
+      "learning_rate": 4.7077980438805185e-05,
+      "loss": 1.3269,
+      "step": 11064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8683897256851196,
+      "learning_rate": 4.707771609833466e-05,
+      "loss": 1.3982,
+      "step": 11065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7584537267684937,
+      "learning_rate": 4.707745175786413e-05,
+      "loss": 1.5186,
+      "step": 11066
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5781164169311523,
+      "learning_rate": 4.7077187417393606e-05,
+      "loss": 1.4589,
+      "step": 11067
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.994931697845459,
+      "learning_rate": 4.707692307692308e-05,
+      "loss": 1.614,
+      "step": 11068
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4951488971710205,
+      "learning_rate": 4.707665873645255e-05,
+      "loss": 1.3,
+      "step": 11069
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.787449836730957,
+      "learning_rate": 4.707639439598203e-05,
+      "loss": 1.4445,
+      "step": 11070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9771475791931152,
+      "learning_rate": 4.70761300555115e-05,
+      "loss": 1.4971,
+      "step": 11071
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2596657276153564,
+      "learning_rate": 4.7075865715040974e-05,
+      "loss": 1.2657,
+      "step": 11072
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6280226707458496,
+      "learning_rate": 4.707560137457045e-05,
+      "loss": 1.3403,
+      "step": 11073
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4588258266448975,
+      "learning_rate": 4.707533703409993e-05,
+      "loss": 1.474,
+      "step": 11074
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171375274658203,
+      "learning_rate": 4.7075072693629395e-05,
+      "loss": 1.4502,
+      "step": 11075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.310943365097046,
+      "learning_rate": 4.7074808353158875e-05,
+      "loss": 1.3015,
+      "step": 11076
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5730479955673218,
+      "learning_rate": 4.707454401268834e-05,
+      "loss": 1.411,
+      "step": 11077
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8846073150634766,
+      "learning_rate": 4.707427967221782e-05,
+      "loss": 1.3919,
+      "step": 11078
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7609636783599854,
+      "learning_rate": 4.7074015331747296e-05,
+      "loss": 1.4848,
+      "step": 11079
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0698301792144775,
+      "learning_rate": 4.707375099127677e-05,
+      "loss": 1.6025,
+      "step": 11080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6268789768218994,
+      "learning_rate": 4.707348665080624e-05,
+      "loss": 1.2669,
+      "step": 11081
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.30145263671875,
+      "learning_rate": 4.707322231033571e-05,
+      "loss": 1.4742,
+      "step": 11082
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.169010877609253,
+      "learning_rate": 4.707295796986519e-05,
+      "loss": 1.4621,
+      "step": 11083
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0834107398986816,
+      "learning_rate": 4.7072693629394664e-05,
+      "loss": 1.4051,
+      "step": 11084
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.47490656375885,
+      "learning_rate": 4.707242928892414e-05,
+      "loss": 1.2928,
+      "step": 11085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8550976514816284,
+      "learning_rate": 4.707216494845361e-05,
+      "loss": 1.4314,
+      "step": 11086
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5562729835510254,
+      "learning_rate": 4.7071900607983084e-05,
+      "loss": 1.3482,
+      "step": 11087
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5568573474884033,
+      "learning_rate": 4.707163626751256e-05,
+      "loss": 1.4499,
+      "step": 11088
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3028109073638916,
+      "learning_rate": 4.707137192704204e-05,
+      "loss": 1.4669,
+      "step": 11089
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9857206344604492,
+      "learning_rate": 4.7071107586571505e-05,
+      "loss": 1.3777,
+      "step": 11090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.04642915725708,
+      "learning_rate": 4.707084324610098e-05,
+      "loss": 1.8675,
+      "step": 11091
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3160157203674316,
+      "learning_rate": 4.707057890563045e-05,
+      "loss": 1.495,
+      "step": 11092
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.428141474723816,
+      "learning_rate": 4.7070314565159926e-05,
+      "loss": 1.2122,
+      "step": 11093
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3660659790039062,
+      "learning_rate": 4.7070050224689406e-05,
+      "loss": 1.4622,
+      "step": 11094
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.204935312271118,
+      "learning_rate": 4.706978588421887e-05,
+      "loss": 1.5389,
+      "step": 11095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.618046522140503,
+      "learning_rate": 4.706952154374835e-05,
+      "loss": 1.3909,
+      "step": 11096
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8939104080200195,
+      "learning_rate": 4.706925720327782e-05,
+      "loss": 1.4316,
+      "step": 11097
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7023612260818481,
+      "learning_rate": 4.70689928628073e-05,
+      "loss": 1.4049,
+      "step": 11098
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3224098682403564,
+      "learning_rate": 4.7068728522336774e-05,
+      "loss": 1.6335,
+      "step": 11099
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1207470893859863,
+      "learning_rate": 4.706846418186625e-05,
+      "loss": 1.3508,
+      "step": 11100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.610753059387207,
+      "learning_rate": 4.706819984139572e-05,
+      "loss": 1.3412,
+      "step": 11101
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.840658187866211,
+      "learning_rate": 4.7067935500925195e-05,
+      "loss": 1.3675,
+      "step": 11102
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.793820858001709,
+      "learning_rate": 4.706767116045467e-05,
+      "loss": 1.3036,
+      "step": 11103
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5553014278411865,
+      "learning_rate": 4.706740681998414e-05,
+      "loss": 1.4919,
+      "step": 11104
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8908429145812988,
+      "learning_rate": 4.7067142479513615e-05,
+      "loss": 1.4963,
+      "step": 11105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.632610559463501,
+      "learning_rate": 4.706687813904309e-05,
+      "loss": 1.3778,
+      "step": 11106
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9503213167190552,
+      "learning_rate": 4.706661379857256e-05,
+      "loss": 1.2165,
+      "step": 11107
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6864935159683228,
+      "learning_rate": 4.7066349458102036e-05,
+      "loss": 1.5113,
+      "step": 11108
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5922925472259521,
+      "learning_rate": 4.7066085117631516e-05,
+      "loss": 1.394,
+      "step": 11109
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5620447397232056,
+      "learning_rate": 4.706582077716098e-05,
+      "loss": 1.4182,
+      "step": 11110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.621809959411621,
+      "learning_rate": 4.7065556436690464e-05,
+      "loss": 1.3337,
+      "step": 11111
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2173547744750977,
+      "learning_rate": 4.706529209621993e-05,
+      "loss": 1.3519,
+      "step": 11112
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5694019794464111,
+      "learning_rate": 4.706502775574941e-05,
+      "loss": 1.3646,
+      "step": 11113
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4783480167388916,
+      "learning_rate": 4.7064763415278884e-05,
+      "loss": 1.3032,
+      "step": 11114
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7696678638458252,
+      "learning_rate": 4.706449907480836e-05,
+      "loss": 1.3975,
+      "step": 11115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.04970383644104,
+      "learning_rate": 4.706423473433783e-05,
+      "loss": 2.3465,
+      "step": 11116
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6536086797714233,
+      "learning_rate": 4.70639703938673e-05,
+      "loss": 1.3342,
+      "step": 11117
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6000407934188843,
+      "learning_rate": 4.706370605339678e-05,
+      "loss": 1.4794,
+      "step": 11118
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8294992446899414,
+      "learning_rate": 4.706344171292625e-05,
+      "loss": 1.4406,
+      "step": 11119
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7798864841461182,
+      "learning_rate": 4.7063177372455726e-05,
+      "loss": 1.4614,
+      "step": 11120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6038254499435425,
+      "learning_rate": 4.70629130319852e-05,
+      "loss": 1.4862,
+      "step": 11121
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6382111310958862,
+      "learning_rate": 4.706264869151467e-05,
+      "loss": 1.4296,
+      "step": 11122
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1698126792907715,
+      "learning_rate": 4.7062384351044146e-05,
+      "loss": 1.6175,
+      "step": 11123
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5257304906845093,
+      "learning_rate": 4.706212001057363e-05,
+      "loss": 1.415,
+      "step": 11124
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3526647090911865,
+      "learning_rate": 4.7061855670103094e-05,
+      "loss": 1.4489,
+      "step": 11125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5004546642303467,
+      "learning_rate": 4.7061591329632574e-05,
+      "loss": 1.3412,
+      "step": 11126
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.861567735671997,
+      "learning_rate": 4.706132698916204e-05,
+      "loss": 1.2442,
+      "step": 11127
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.813722848892212,
+      "learning_rate": 4.7061062648691514e-05,
+      "loss": 1.3263,
+      "step": 11128
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.622527837753296,
+      "learning_rate": 4.7060798308220995e-05,
+      "loss": 1.3503,
+      "step": 11129
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6744253635406494,
+      "learning_rate": 4.706053396775046e-05,
+      "loss": 1.3695,
+      "step": 11130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.881418228149414,
+      "learning_rate": 4.706026962727994e-05,
+      "loss": 1.453,
+      "step": 11131
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.854164481163025,
+      "learning_rate": 4.706000528680941e-05,
+      "loss": 1.4696,
+      "step": 11132
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9267122745513916,
+      "learning_rate": 4.705974094633889e-05,
+      "loss": 1.3976,
+      "step": 11133
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.588976502418518,
+      "learning_rate": 4.705947660586836e-05,
+      "loss": 1.4665,
+      "step": 11134
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.230851173400879,
+      "learning_rate": 4.7059212265397836e-05,
+      "loss": 1.48,
+      "step": 11135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.027073621749878,
+      "learning_rate": 4.705894792492731e-05,
+      "loss": 1.5344,
+      "step": 11136
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5223289728164673,
+      "learning_rate": 4.705868358445678e-05,
+      "loss": 1.9086,
+      "step": 11137
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4056504964828491,
+      "learning_rate": 4.705841924398626e-05,
+      "loss": 2.074,
+      "step": 11138
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.222804307937622,
+      "learning_rate": 4.705815490351573e-05,
+      "loss": 1.4149,
+      "step": 11139
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5670181512832642,
+      "learning_rate": 4.7057890563045204e-05,
+      "loss": 1.3102,
+      "step": 11140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9730931520462036,
+      "learning_rate": 4.705762622257468e-05,
+      "loss": 1.4617,
+      "step": 11141
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8496087789535522,
+      "learning_rate": 4.705736188210415e-05,
+      "loss": 1.4068,
+      "step": 11142
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4455406665802002,
+      "learning_rate": 4.7057097541633625e-05,
+      "loss": 1.2504,
+      "step": 11143
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6128402948379517,
+      "learning_rate": 4.7056833201163105e-05,
+      "loss": 1.3213,
+      "step": 11144
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7200357913970947,
+      "learning_rate": 4.705656886069257e-05,
+      "loss": 1.3281,
+      "step": 11145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6022943258285522,
+      "learning_rate": 4.705630452022205e-05,
+      "loss": 1.3801,
+      "step": 11146
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0117299556732178,
+      "learning_rate": 4.705604017975152e-05,
+      "loss": 1.5338,
+      "step": 11147
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4771406650543213,
+      "learning_rate": 4.7055775839281e-05,
+      "loss": 1.3711,
+      "step": 11148
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.062847137451172,
+      "learning_rate": 4.705551149881047e-05,
+      "loss": 1.5042,
+      "step": 11149
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.491956949234009,
+      "learning_rate": 4.7055247158339946e-05,
+      "loss": 1.4675,
+      "step": 11150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.433411717414856,
+      "learning_rate": 4.705498281786942e-05,
+      "loss": 1.2991,
+      "step": 11151
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7043530941009521,
+      "learning_rate": 4.7054718477398894e-05,
+      "loss": 1.4495,
+      "step": 11152
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4896976947784424,
+      "learning_rate": 4.705445413692837e-05,
+      "loss": 1.4441,
+      "step": 11153
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4572596549987793,
+      "learning_rate": 4.705418979645784e-05,
+      "loss": 1.5709,
+      "step": 11154
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9025428295135498,
+      "learning_rate": 4.7053925455987314e-05,
+      "loss": 1.3671,
+      "step": 11155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4837119579315186,
+      "learning_rate": 4.705366111551679e-05,
+      "loss": 1.3454,
+      "step": 11156
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5563843250274658,
+      "learning_rate": 4.705339677504626e-05,
+      "loss": 1.2269,
+      "step": 11157
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.002715826034546,
+      "learning_rate": 4.7053132434575735e-05,
+      "loss": 1.4953,
+      "step": 11158
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6298346519470215,
+      "learning_rate": 4.7052868094105215e-05,
+      "loss": 1.7871,
+      "step": 11159
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9901299476623535,
+      "learning_rate": 4.705260375363468e-05,
+      "loss": 1.5946,
+      "step": 11160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3989195823669434,
+      "learning_rate": 4.705233941316416e-05,
+      "loss": 1.6216,
+      "step": 11161
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8260109424591064,
+      "learning_rate": 4.705207507269363e-05,
+      "loss": 1.4297,
+      "step": 11162
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.222247362136841,
+      "learning_rate": 4.70518107322231e-05,
+      "loss": 1.5404,
+      "step": 11163
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.111398935317993,
+      "learning_rate": 4.705154639175258e-05,
+      "loss": 1.5374,
+      "step": 11164
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9548403024673462,
+      "learning_rate": 4.705128205128205e-05,
+      "loss": 1.5606,
+      "step": 11165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0808048248291016,
+      "learning_rate": 4.705101771081153e-05,
+      "loss": 1.3629,
+      "step": 11166
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7143113613128662,
+      "learning_rate": 4.7050753370341e-05,
+      "loss": 1.4083,
+      "step": 11167
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9158293008804321,
+      "learning_rate": 4.705048902987048e-05,
+      "loss": 1.3416,
+      "step": 11168
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2241194248199463,
+      "learning_rate": 4.705022468939995e-05,
+      "loss": 1.5583,
+      "step": 11169
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9357779026031494,
+      "learning_rate": 4.7049960348929425e-05,
+      "loss": 1.2381,
+      "step": 11170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8463554382324219,
+      "learning_rate": 4.70496960084589e-05,
+      "loss": 1.337,
+      "step": 11171
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.191152572631836,
+      "learning_rate": 4.704943166798837e-05,
+      "loss": 1.511,
+      "step": 11172
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7034478187561035,
+      "learning_rate": 4.7049167327517845e-05,
+      "loss": 1.3483,
+      "step": 11173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.829918384552002,
+      "learning_rate": 4.704890298704732e-05,
+      "loss": 1.5746,
+      "step": 11174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6673340797424316,
+      "learning_rate": 4.704863864657679e-05,
+      "loss": 1.3177,
+      "step": 11175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5017507076263428,
+      "learning_rate": 4.7048374306106266e-05,
+      "loss": 1.5886,
+      "step": 11176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0093491077423096,
+      "learning_rate": 4.704810996563574e-05,
+      "loss": 1.4401,
+      "step": 11177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.458000898361206,
+      "learning_rate": 4.704784562516521e-05,
+      "loss": 1.2441,
+      "step": 11178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1042397022247314,
+      "learning_rate": 4.7047581284694693e-05,
+      "loss": 1.4686,
+      "step": 11179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.620327353477478,
+      "learning_rate": 4.704731694422416e-05,
+      "loss": 1.3995,
+      "step": 11180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1028590202331543,
+      "learning_rate": 4.704705260375364e-05,
+      "loss": 1.3712,
+      "step": 11181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5521373748779297,
+      "learning_rate": 4.704678826328311e-05,
+      "loss": 1.2605,
+      "step": 11182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8442108631134033,
+      "learning_rate": 4.704652392281259e-05,
+      "loss": 1.5192,
+      "step": 11183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3239703178405762,
+      "learning_rate": 4.704625958234206e-05,
+      "loss": 1.9019,
+      "step": 11184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5493323802947998,
+      "learning_rate": 4.7045995241871535e-05,
+      "loss": 1.3018,
+      "step": 11185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2906126976013184,
+      "learning_rate": 4.704573090140101e-05,
+      "loss": 1.4783,
+      "step": 11186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1507349014282227,
+      "learning_rate": 4.704546656093048e-05,
+      "loss": 1.5498,
+      "step": 11187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7951725721359253,
+      "learning_rate": 4.7045202220459956e-05,
+      "loss": 1.4353,
+      "step": 11188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09565806388855,
+      "learning_rate": 4.704493787998943e-05,
+      "loss": 1.4238,
+      "step": 11189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6461317539215088,
+      "learning_rate": 4.70446735395189e-05,
+      "loss": 1.4008,
+      "step": 11190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7365726232528687,
+      "learning_rate": 4.7044409199048376e-05,
+      "loss": 1.8708,
+      "step": 11191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7764278650283813,
+      "learning_rate": 4.704414485857785e-05,
+      "loss": 1.3817,
+      "step": 11192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8335015773773193,
+      "learning_rate": 4.7043880518107323e-05,
+      "loss": 1.2177,
+      "step": 11193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7543466091156006,
+      "learning_rate": 4.7043616177636804e-05,
+      "loss": 1.2922,
+      "step": 11194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5542066097259521,
+      "learning_rate": 4.704335183716627e-05,
+      "loss": 1.1609,
+      "step": 11195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.685975432395935,
+      "learning_rate": 4.704308749669575e-05,
+      "loss": 1.2007,
+      "step": 11196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7658294439315796,
+      "learning_rate": 4.704282315622522e-05,
+      "loss": 1.4675,
+      "step": 11197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8913543224334717,
+      "learning_rate": 4.70425588157547e-05,
+      "loss": 1.4336,
+      "step": 11198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6081316471099854,
+      "learning_rate": 4.704229447528417e-05,
+      "loss": 1.3923,
+      "step": 11199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6886463165283203,
+      "learning_rate": 4.704203013481364e-05,
+      "loss": 1.469,
+      "step": 11200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4674649238586426,
+      "learning_rate": 4.704176579434312e-05,
+      "loss": 1.2833,
+      "step": 11201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9669560194015503,
+      "learning_rate": 4.7041501453872586e-05,
+      "loss": 1.3655,
+      "step": 11202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3466782569885254,
+      "learning_rate": 4.7041237113402066e-05,
+      "loss": 1.5027,
+      "step": 11203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9145230054855347,
+      "learning_rate": 4.704097277293154e-05,
+      "loss": 1.5607,
+      "step": 11204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5585136413574219,
+      "learning_rate": 4.704070843246101e-05,
+      "loss": 1.2723,
+      "step": 11205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6525970697402954,
+      "learning_rate": 4.704044409199049e-05,
+      "loss": 1.4349,
+      "step": 11206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.195240020751953,
+      "learning_rate": 4.704017975151996e-05,
+      "loss": 1.6051,
+      "step": 11207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5657055377960205,
+      "learning_rate": 4.7039915411049434e-05,
+      "loss": 1.3279,
+      "step": 11208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8335316181182861,
+      "learning_rate": 4.703965107057891e-05,
+      "loss": 1.3101,
+      "step": 11209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5304726362228394,
+      "learning_rate": 4.703938673010838e-05,
+      "loss": 1.3153,
+      "step": 11210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9444358348846436,
+      "learning_rate": 4.7039122389637855e-05,
+      "loss": 1.4481,
+      "step": 11211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4487287998199463,
+      "learning_rate": 4.703885804916733e-05,
+      "loss": 1.6961,
+      "step": 11212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.651781678199768,
+      "learning_rate": 4.70385937086968e-05,
+      "loss": 1.303,
+      "step": 11213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6153111457824707,
+      "learning_rate": 4.703832936822628e-05,
+      "loss": 1.2917,
+      "step": 11214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.425144910812378,
+      "learning_rate": 4.703806502775575e-05,
+      "loss": 1.4177,
+      "step": 11215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3872029781341553,
+      "learning_rate": 4.703780068728523e-05,
+      "loss": 1.4937,
+      "step": 11216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2712643146514893,
+      "learning_rate": 4.7037536346814696e-05,
+      "loss": 1.4678,
+      "step": 11217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7730844020843506,
+      "learning_rate": 4.7037272006344176e-05,
+      "loss": 1.2622,
+      "step": 11218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.577538013458252,
+      "learning_rate": 4.703700766587365e-05,
+      "loss": 1.4646,
+      "step": 11219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6661003828048706,
+      "learning_rate": 4.7036743325403123e-05,
+      "loss": 1.5041,
+      "step": 11220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7310311794281006,
+      "learning_rate": 4.70364789849326e-05,
+      "loss": 1.5432,
+      "step": 11221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7375926971435547,
+      "learning_rate": 4.703621464446207e-05,
+      "loss": 1.3899,
+      "step": 11222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4581505060195923,
+      "learning_rate": 4.7035950303991544e-05,
+      "loss": 1.3022,
+      "step": 11223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0395495891571045,
+      "learning_rate": 4.703568596352102e-05,
+      "loss": 1.5164,
+      "step": 11224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.325366735458374,
+      "learning_rate": 4.703542162305049e-05,
+      "loss": 1.5806,
+      "step": 11225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.794505000114441,
+      "learning_rate": 4.7035157282579965e-05,
+      "loss": 1.5268,
+      "step": 11226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.617600440979004,
+      "learning_rate": 4.703489294210944e-05,
+      "loss": 1.3591,
+      "step": 11227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.192199468612671,
+      "learning_rate": 4.703462860163891e-05,
+      "loss": 1.4806,
+      "step": 11228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5759330987930298,
+      "learning_rate": 4.703436426116839e-05,
+      "loss": 1.3924,
+      "step": 11229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234531879425049,
+      "learning_rate": 4.703409992069786e-05,
+      "loss": 1.5146,
+      "step": 11230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5718411207199097,
+      "learning_rate": 4.703383558022734e-05,
+      "loss": 1.236,
+      "step": 11231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6681567430496216,
+      "learning_rate": 4.7033571239756806e-05,
+      "loss": 1.4064,
+      "step": 11232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.80014967918396,
+      "learning_rate": 4.703330689928629e-05,
+      "loss": 1.1536,
+      "step": 11233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7029694318771362,
+      "learning_rate": 4.703304255881576e-05,
+      "loss": 1.3414,
+      "step": 11234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5888562202453613,
+      "learning_rate": 4.703277821834523e-05,
+      "loss": 1.4567,
+      "step": 11235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4436489343643188,
+      "learning_rate": 4.703251387787471e-05,
+      "loss": 1.2516,
+      "step": 11236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.128107786178589,
+      "learning_rate": 4.7032249537404174e-05,
+      "loss": 1.3851,
+      "step": 11237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.030038356781006,
+      "learning_rate": 4.7031985196933654e-05,
+      "loss": 1.4819,
+      "step": 11238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8281255960464478,
+      "learning_rate": 4.703172085646313e-05,
+      "loss": 1.4117,
+      "step": 11239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5573195219039917,
+      "learning_rate": 4.70314565159926e-05,
+      "loss": 1.3075,
+      "step": 11240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5740361213684082,
+      "learning_rate": 4.7031192175522075e-05,
+      "loss": 1.3084,
+      "step": 11241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2661616802215576,
+      "learning_rate": 4.703092783505155e-05,
+      "loss": 1.4074,
+      "step": 11242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2291667461395264,
+      "learning_rate": 4.703066349458102e-05,
+      "loss": 1.3799,
+      "step": 11243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8058111667633057,
+      "learning_rate": 4.7030399154110496e-05,
+      "loss": 1.2284,
+      "step": 11244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6383846998214722,
+      "learning_rate": 4.703013481363997e-05,
+      "loss": 1.3358,
+      "step": 11245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7240773439407349,
+      "learning_rate": 4.702987047316944e-05,
+      "loss": 1.3933,
+      "step": 11246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5311936140060425,
+      "learning_rate": 4.702960613269892e-05,
+      "loss": 1.3674,
+      "step": 11247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6476609706878662,
+      "learning_rate": 4.702934179222839e-05,
+      "loss": 1.3456,
+      "step": 11248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3939316272735596,
+      "learning_rate": 4.7029077451757864e-05,
+      "loss": 1.3338,
+      "step": 11249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8621032238006592,
+      "learning_rate": 4.702881311128734e-05,
+      "loss": 1.451,
+      "step": 11250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.710431694984436,
+      "learning_rate": 4.702854877081682e-05,
+      "loss": 1.3681,
+      "step": 11251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8888213634490967,
+      "learning_rate": 4.7028284430346284e-05,
+      "loss": 1.6438,
+      "step": 11252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9655299186706543,
+      "learning_rate": 4.7028020089875765e-05,
+      "loss": 1.3371,
+      "step": 11253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7146320343017578,
+      "learning_rate": 4.702775574940523e-05,
+      "loss": 1.4747,
+      "step": 11254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3711917400360107,
+      "learning_rate": 4.702749140893471e-05,
+      "loss": 1.437,
+      "step": 11255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.197073221206665,
+      "learning_rate": 4.7027227068464186e-05,
+      "loss": 1.3377,
+      "step": 11256
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6773909330368042,
+      "learning_rate": 4.702696272799366e-05,
+      "loss": 1.2654,
+      "step": 11257
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3982548713684082,
+      "learning_rate": 4.702669838752313e-05,
+      "loss": 1.3117,
+      "step": 11258
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2533578872680664,
+      "learning_rate": 4.7026434047052606e-05,
+      "loss": 1.5232,
+      "step": 11259
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8905208110809326,
+      "learning_rate": 4.702616970658208e-05,
+      "loss": 1.4084,
+      "step": 11260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9168492555618286,
+      "learning_rate": 4.7025905366111553e-05,
+      "loss": 1.4686,
+      "step": 11261
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0575218200683594,
+      "learning_rate": 4.702564102564103e-05,
+      "loss": 1.464,
+      "step": 11262
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6451441049575806,
+      "learning_rate": 4.70253766851705e-05,
+      "loss": 1.4259,
+      "step": 11263
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.018270492553711,
+      "learning_rate": 4.7025112344699974e-05,
+      "loss": 1.3914,
+      "step": 11264
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.958569049835205,
+      "learning_rate": 4.702484800422945e-05,
+      "loss": 1.4188,
+      "step": 11265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7164666652679443,
+      "learning_rate": 4.702458366375893e-05,
+      "loss": 1.4152,
+      "step": 11266
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3948835134506226,
+      "learning_rate": 4.7024319323288395e-05,
+      "loss": 1.5929,
+      "step": 11267
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2298402786254883,
+      "learning_rate": 4.7024054982817875e-05,
+      "loss": 1.4546,
+      "step": 11268
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1512351036071777,
+      "learning_rate": 4.702379064234734e-05,
+      "loss": 1.1861,
+      "step": 11269
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5059747695922852,
+      "learning_rate": 4.702352630187682e-05,
+      "loss": 1.3135,
+      "step": 11270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4464555978775024,
+      "learning_rate": 4.7023261961406296e-05,
+      "loss": 1.2896,
+      "step": 11271
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7490007877349854,
+      "learning_rate": 4.702299762093576e-05,
+      "loss": 1.7435,
+      "step": 11272
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.034594774246216,
+      "learning_rate": 4.702273328046524e-05,
+      "loss": 1.487,
+      "step": 11273
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.586371898651123,
+      "learning_rate": 4.702246893999471e-05,
+      "loss": 1.3437,
+      "step": 11274
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0379393100738525,
+      "learning_rate": 4.702220459952419e-05,
+      "loss": 1.4561,
+      "step": 11275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.530497431755066,
+      "learning_rate": 4.7021940259053664e-05,
+      "loss": 1.2658,
+      "step": 11276
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.817766785621643,
+      "learning_rate": 4.702167591858314e-05,
+      "loss": 1.307,
+      "step": 11277
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0770492553710938,
+      "learning_rate": 4.702141157811261e-05,
+      "loss": 1.4227,
+      "step": 11278
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.633330225944519,
+      "learning_rate": 4.7021147237642084e-05,
+      "loss": 1.3501,
+      "step": 11279
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.848327875137329,
+      "learning_rate": 4.702088289717156e-05,
+      "loss": 1.3718,
+      "step": 11280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0928966999053955,
+      "learning_rate": 4.702061855670103e-05,
+      "loss": 1.5687,
+      "step": 11281
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5625698566436768,
+      "learning_rate": 4.7020354216230505e-05,
+      "loss": 1.3804,
+      "step": 11282
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2093968391418457,
+      "learning_rate": 4.702008987575998e-05,
+      "loss": 1.4712,
+      "step": 11283
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8393902778625488,
+      "learning_rate": 4.701982553528945e-05,
+      "loss": 1.2585,
+      "step": 11284
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2144689559936523,
+      "learning_rate": 4.7019561194818926e-05,
+      "loss": 1.3622,
+      "step": 11285
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.454689860343933,
+      "learning_rate": 4.7019296854348406e-05,
+      "loss": 1.3004,
+      "step": 11286
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.417535662651062,
+      "learning_rate": 4.701903251387787e-05,
+      "loss": 1.2831,
+      "step": 11287
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8715291023254395,
+      "learning_rate": 4.701876817340735e-05,
+      "loss": 1.4864,
+      "step": 11288
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.933671236038208,
+      "learning_rate": 4.701850383293682e-05,
+      "loss": 1.5502,
+      "step": 11289
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7325315475463867,
+      "learning_rate": 4.70182394924663e-05,
+      "loss": 1.5105,
+      "step": 11290
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.702040433883667,
+      "learning_rate": 4.7017975151995774e-05,
+      "loss": 1.4391,
+      "step": 11291
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9307167530059814,
+      "learning_rate": 4.701771081152525e-05,
+      "loss": 1.5259,
+      "step": 11292
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.144676923751831,
+      "learning_rate": 4.701744647105472e-05,
+      "loss": 1.3738,
+      "step": 11293
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.388454556465149,
+      "learning_rate": 4.7017182130584195e-05,
+      "loss": 1.3439,
+      "step": 11294
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.56003999710083,
+      "learning_rate": 4.701691779011367e-05,
+      "loss": 1.3474,
+      "step": 11295
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6823688745498657,
+      "learning_rate": 4.701665344964314e-05,
+      "loss": 1.4067,
+      "step": 11296
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.350910186767578,
+      "learning_rate": 4.7016389109172616e-05,
+      "loss": 1.5929,
+      "step": 11297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.593127727508545,
+      "learning_rate": 4.701612476870209e-05,
+      "loss": 1.3566,
+      "step": 11298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.032620668411255,
+      "learning_rate": 4.701586042823156e-05,
+      "loss": 1.4585,
+      "step": 11299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8683403730392456,
+      "learning_rate": 4.7015596087761036e-05,
+      "loss": 1.3929,
+      "step": 11300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.68630051612854,
+      "learning_rate": 4.7015331747290517e-05,
+      "loss": 1.3621,
+      "step": 11301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.016232967376709,
+      "learning_rate": 4.701506740681998e-05,
+      "loss": 1.2965,
+      "step": 11302
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6882483959197998,
+      "learning_rate": 4.7014803066349464e-05,
+      "loss": 1.4777,
+      "step": 11303
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.847066044807434,
+      "learning_rate": 4.701453872587893e-05,
+      "loss": 1.4706,
+      "step": 11304
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.188258647918701,
+      "learning_rate": 4.701427438540841e-05,
+      "loss": 1.3781,
+      "step": 11305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.4283013343811035,
+      "learning_rate": 4.7014010044937884e-05,
+      "loss": 1.5555,
+      "step": 11306
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265618324279785,
+      "learning_rate": 4.701374570446735e-05,
+      "loss": 1.5691,
+      "step": 11307
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8163344860076904,
+      "learning_rate": 4.701348136399683e-05,
+      "loss": 1.3987,
+      "step": 11308
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5255966186523438,
+      "learning_rate": 4.70132170235263e-05,
+      "loss": 1.2949,
+      "step": 11309
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5638567209243774,
+      "learning_rate": 4.701295268305578e-05,
+      "loss": 1.421,
+      "step": 11310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4858819246292114,
+      "learning_rate": 4.701268834258525e-05,
+      "loss": 1.1399,
+      "step": 11311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0046045780181885,
+      "learning_rate": 4.7012424002114726e-05,
+      "loss": 1.6249,
+      "step": 11312
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8342212438583374,
+      "learning_rate": 4.70121596616442e-05,
+      "loss": 1.4473,
+      "step": 11313
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4912654161453247,
+      "learning_rate": 4.701189532117367e-05,
+      "loss": 1.2527,
+      "step": 11314
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8286949396133423,
+      "learning_rate": 4.7011630980703147e-05,
+      "loss": 1.251,
+      "step": 11315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7107157707214355,
+      "learning_rate": 4.701136664023263e-05,
+      "loss": 1.4406,
+      "step": 11316
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6528127193450928,
+      "learning_rate": 4.7011102299762094e-05,
+      "loss": 1.3065,
+      "step": 11317
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.490036129951477,
+      "learning_rate": 4.701083795929157e-05,
+      "loss": 1.313,
+      "step": 11318
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.815813422203064,
+      "learning_rate": 4.701057361882104e-05,
+      "loss": 1.4465,
+      "step": 11319
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.73106050491333,
+      "learning_rate": 4.7010309278350514e-05,
+      "loss": 1.4046,
+      "step": 11320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9626070261001587,
+      "learning_rate": 4.7010044937879995e-05,
+      "loss": 1.5816,
+      "step": 11321
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.145875930786133,
+      "learning_rate": 4.700978059740946e-05,
+      "loss": 1.3509,
+      "step": 11322
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.905884027481079,
+      "learning_rate": 4.700951625693894e-05,
+      "loss": 1.516,
+      "step": 11323
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.917957067489624,
+      "learning_rate": 4.700925191646841e-05,
+      "loss": 1.5107,
+      "step": 11324
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1069021224975586,
+      "learning_rate": 4.700898757599789e-05,
+      "loss": 1.58,
+      "step": 11325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8814446926116943,
+      "learning_rate": 4.700872323552736e-05,
+      "loss": 1.2631,
+      "step": 11326
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4806164503097534,
+      "learning_rate": 4.7008458895056836e-05,
+      "loss": 1.3075,
+      "step": 11327
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3518040180206299,
+      "learning_rate": 4.700819455458631e-05,
+      "loss": 1.3626,
+      "step": 11328
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0995657444000244,
+      "learning_rate": 4.700793021411578e-05,
+      "loss": 1.3721,
+      "step": 11329
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4558768272399902,
+      "learning_rate": 4.700766587364526e-05,
+      "loss": 1.2302,
+      "step": 11330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.373252511024475,
+      "learning_rate": 4.700740153317473e-05,
+      "loss": 1.185,
+      "step": 11331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9724650382995605,
+      "learning_rate": 4.7007137192704204e-05,
+      "loss": 1.6753,
+      "step": 11332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5445420742034912,
+      "learning_rate": 4.700687285223368e-05,
+      "loss": 1.2611,
+      "step": 11333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9706848859786987,
+      "learning_rate": 4.700660851176315e-05,
+      "loss": 1.4241,
+      "step": 11334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0870847702026367,
+      "learning_rate": 4.7006344171292625e-05,
+      "loss": 1.4592,
+      "step": 11335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5911883115768433,
+      "learning_rate": 4.7006079830822105e-05,
+      "loss": 1.3654,
+      "step": 11336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4356964826583862,
+      "learning_rate": 4.700581549035157e-05,
+      "loss": 1.3449,
+      "step": 11337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.476717233657837,
+      "learning_rate": 4.700555114988105e-05,
+      "loss": 1.3332,
+      "step": 11338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5971492528915405,
+      "learning_rate": 4.700528680941052e-05,
+      "loss": 1.2108,
+      "step": 11339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3981244564056396,
+      "learning_rate": 4.700502246894e-05,
+      "loss": 1.2912,
+      "step": 11340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7446765899658203,
+      "learning_rate": 4.700475812846947e-05,
+      "loss": 1.3274,
+      "step": 11341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.168104887008667,
+      "learning_rate": 4.7004493787998947e-05,
+      "loss": 1.48,
+      "step": 11342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7940597534179688,
+      "learning_rate": 4.700422944752842e-05,
+      "loss": 1.4242,
+      "step": 11343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.612725019454956,
+      "learning_rate": 4.700396510705789e-05,
+      "loss": 1.3682,
+      "step": 11344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0900416374206543,
+      "learning_rate": 4.700370076658737e-05,
+      "loss": 1.5604,
+      "step": 11345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.439960479736328,
+      "learning_rate": 4.700343642611684e-05,
+      "loss": 1.4352,
+      "step": 11346
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.512626051902771,
+      "learning_rate": 4.7003172085646314e-05,
+      "loss": 1.122,
+      "step": 11347
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.968075156211853,
+      "learning_rate": 4.700290774517579e-05,
+      "loss": 1.4247,
+      "step": 11348
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8382803201675415,
+      "learning_rate": 4.700264340470526e-05,
+      "loss": 1.4138,
+      "step": 11349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046894073486328,
+      "learning_rate": 4.7002379064234735e-05,
+      "loss": 1.4847,
+      "step": 11350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7077696323394775,
+      "learning_rate": 4.7002114723764215e-05,
+      "loss": 1.4235,
+      "step": 11351
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.415858507156372,
+      "learning_rate": 4.700185038329368e-05,
+      "loss": 1.5825,
+      "step": 11352
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1157805919647217,
+      "learning_rate": 4.7001586042823156e-05,
+      "loss": 1.4104,
+      "step": 11353
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.189138412475586,
+      "learning_rate": 4.700132170235263e-05,
+      "loss": 1.4576,
+      "step": 11354
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2382404804229736,
+      "learning_rate": 4.70010573618821e-05,
+      "loss": 1.3819,
+      "step": 11355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8543965816497803,
+      "learning_rate": 4.700079302141158e-05,
+      "loss": 1.4605,
+      "step": 11356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.545688271522522,
+      "learning_rate": 4.700052868094105e-05,
+      "loss": 1.3683,
+      "step": 11357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5473225116729736,
+      "learning_rate": 4.700026434047053e-05,
+      "loss": 1.2235,
+      "step": 11358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8544625043869019,
+      "learning_rate": 4.7e-05,
+      "loss": 1.2071,
+      "step": 11359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9714343547821045,
+      "learning_rate": 4.699973565952948e-05,
+      "loss": 1.3338,
+      "step": 11360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.108016014099121,
+      "learning_rate": 4.699947131905895e-05,
+      "loss": 1.3283,
+      "step": 11361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4072319269180298,
+      "learning_rate": 4.6999206978588425e-05,
+      "loss": 1.425,
+      "step": 11362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.406226396560669,
+      "learning_rate": 4.69989426381179e-05,
+      "loss": 1.3197,
+      "step": 11363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8170133829116821,
+      "learning_rate": 4.699867829764737e-05,
+      "loss": 1.383,
+      "step": 11364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5211060047149658,
+      "learning_rate": 4.6998413957176845e-05,
+      "loss": 1.3329,
+      "step": 11365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5442330837249756,
+      "learning_rate": 4.699814961670632e-05,
+      "loss": 1.3053,
+      "step": 11366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8775006532669067,
+      "learning_rate": 4.699788527623579e-05,
+      "loss": 1.4506,
+      "step": 11367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0282413959503174,
+      "learning_rate": 4.6997620935765266e-05,
+      "loss": 1.2688,
+      "step": 11368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.823037028312683,
+      "learning_rate": 4.699735659529474e-05,
+      "loss": 1.4271,
+      "step": 11369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7210907936096191,
+      "learning_rate": 4.699709225482421e-05,
+      "loss": 1.3393,
+      "step": 11370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3154138326644897,
+      "learning_rate": 4.6996827914353694e-05,
+      "loss": 2.003,
+      "step": 11371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.222952365875244,
+      "learning_rate": 4.699656357388316e-05,
+      "loss": 1.4204,
+      "step": 11372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1022114753723145,
+      "learning_rate": 4.699629923341264e-05,
+      "loss": 1.5049,
+      "step": 11373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6423346996307373,
+      "learning_rate": 4.699603489294211e-05,
+      "loss": 1.4379,
+      "step": 11374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.061591148376465,
+      "learning_rate": 4.699577055247159e-05,
+      "loss": 1.5381,
+      "step": 11375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3682949542999268,
+      "learning_rate": 4.699550621200106e-05,
+      "loss": 1.2005,
+      "step": 11376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.477034568786621,
+      "learning_rate": 4.6995241871530535e-05,
+      "loss": 1.3862,
+      "step": 11377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6712874174118042,
+      "learning_rate": 4.699497753106001e-05,
+      "loss": 1.4186,
+      "step": 11378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.773808717727661,
+      "learning_rate": 4.699471319058948e-05,
+      "loss": 1.5102,
+      "step": 11379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3305304050445557,
+      "learning_rate": 4.6994448850118956e-05,
+      "loss": 1.4461,
+      "step": 11380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7389355897903442,
+      "learning_rate": 4.699418450964843e-05,
+      "loss": 1.4554,
+      "step": 11381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8602213859558105,
+      "learning_rate": 4.69939201691779e-05,
+      "loss": 1.596,
+      "step": 11382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.700879454612732,
+      "learning_rate": 4.6993655828707376e-05,
+      "loss": 1.5496,
+      "step": 11383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.073282480239868,
+      "learning_rate": 4.699339148823685e-05,
+      "loss": 1.3706,
+      "step": 11384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.091155529022217,
+      "learning_rate": 4.6993127147766324e-05,
+      "loss": 1.3577,
+      "step": 11385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3206491470336914,
+      "learning_rate": 4.6992862807295804e-05,
+      "loss": 1.4435,
+      "step": 11386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6594096422195435,
+      "learning_rate": 4.699259846682527e-05,
+      "loss": 1.3472,
+      "step": 11387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5798813104629517,
+      "learning_rate": 4.699233412635475e-05,
+      "loss": 1.4459,
+      "step": 11388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6722766160964966,
+      "learning_rate": 4.699206978588422e-05,
+      "loss": 1.3509,
+      "step": 11389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.099923610687256,
+      "learning_rate": 4.699180544541369e-05,
+      "loss": 1.4042,
+      "step": 11390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390852689743042,
+      "learning_rate": 4.699154110494317e-05,
+      "loss": 1.5916,
+      "step": 11391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7952227592468262,
+      "learning_rate": 4.699127676447264e-05,
+      "loss": 1.366,
+      "step": 11392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6486737728118896,
+      "learning_rate": 4.699101242400212e-05,
+      "loss": 1.2999,
+      "step": 11393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.545398473739624,
+      "learning_rate": 4.6990748083531586e-05,
+      "loss": 1.6497,
+      "step": 11394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8083746433258057,
+      "learning_rate": 4.6990483743061066e-05,
+      "loss": 1.5315,
+      "step": 11395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.210620164871216,
+      "learning_rate": 4.699021940259054e-05,
+      "loss": 1.61,
+      "step": 11396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.650316834449768,
+      "learning_rate": 4.698995506212001e-05,
+      "loss": 1.4119,
+      "step": 11397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.208085775375366,
+      "learning_rate": 4.698969072164949e-05,
+      "loss": 1.5224,
+      "step": 11398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.801588773727417,
+      "learning_rate": 4.698942638117896e-05,
+      "loss": 1.3458,
+      "step": 11399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.567148208618164,
+      "learning_rate": 4.6989162040708434e-05,
+      "loss": 1.1708,
+      "step": 11400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3222613334655762,
+      "learning_rate": 4.698889770023791e-05,
+      "loss": 1.3458,
+      "step": 11401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5909628868103027,
+      "learning_rate": 4.698863335976738e-05,
+      "loss": 1.2915,
+      "step": 11402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1102001667022705,
+      "learning_rate": 4.6988369019296855e-05,
+      "loss": 1.5232,
+      "step": 11403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0260632038116455,
+      "learning_rate": 4.698810467882633e-05,
+      "loss": 1.4798,
+      "step": 11404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7534396648406982,
+      "learning_rate": 4.69878403383558e-05,
+      "loss": 1.4802,
+      "step": 11405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3832067251205444,
+      "learning_rate": 4.698757599788528e-05,
+      "loss": 1.2184,
+      "step": 11406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.565928339958191,
+      "learning_rate": 4.698731165741475e-05,
+      "loss": 1.2096,
+      "step": 11407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0712838172912598,
+      "learning_rate": 4.698704731694423e-05,
+      "loss": 1.3548,
+      "step": 11408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.731917381286621,
+      "learning_rate": 4.6986782976473696e-05,
+      "loss": 1.3944,
+      "step": 11409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5765231847763062,
+      "learning_rate": 4.6986518636003176e-05,
+      "loss": 1.792,
+      "step": 11410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8507851362228394,
+      "learning_rate": 4.698625429553265e-05,
+      "loss": 1.4203,
+      "step": 11411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.592223048210144,
+      "learning_rate": 4.6985989955062124e-05,
+      "loss": 1.2892,
+      "step": 11412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6183772087097168,
+      "learning_rate": 4.69857256145916e-05,
+      "loss": 1.1909,
+      "step": 11413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5370962619781494,
+      "learning_rate": 4.698546127412107e-05,
+      "loss": 1.3568,
+      "step": 11414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5950592756271362,
+      "learning_rate": 4.6985196933650544e-05,
+      "loss": 2.43,
+      "step": 11415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.530063509941101,
+      "learning_rate": 4.698493259318002e-05,
+      "loss": 1.2032,
+      "step": 11416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7462878227233887,
+      "learning_rate": 4.698466825270949e-05,
+      "loss": 1.5073,
+      "step": 11417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2275514602661133,
+      "learning_rate": 4.6984403912238965e-05,
+      "loss": 1.6864,
+      "step": 11418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8290315866470337,
+      "learning_rate": 4.698413957176844e-05,
+      "loss": 1.4239,
+      "step": 11419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1108267307281494,
+      "learning_rate": 4.698387523129791e-05,
+      "loss": 1.4404,
+      "step": 11420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.058692455291748,
+      "learning_rate": 4.698361089082739e-05,
+      "loss": 1.4075,
+      "step": 11421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8439477682113647,
+      "learning_rate": 4.698334655035686e-05,
+      "loss": 1.4459,
+      "step": 11422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6733330488204956,
+      "learning_rate": 4.698308220988634e-05,
+      "loss": 1.4998,
+      "step": 11423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4422874450683594,
+      "learning_rate": 4.6982817869415806e-05,
+      "loss": 1.251,
+      "step": 11424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7081652879714966,
+      "learning_rate": 4.698255352894528e-05,
+      "loss": 1.4465,
+      "step": 11425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8183194398880005,
+      "learning_rate": 4.698228918847476e-05,
+      "loss": 1.3968,
+      "step": 11426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5491456985473633,
+      "learning_rate": 4.698202484800423e-05,
+      "loss": 1.5776,
+      "step": 11427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.061613082885742,
+      "learning_rate": 4.698176050753371e-05,
+      "loss": 1.4207,
+      "step": 11428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.842423677444458,
+      "learning_rate": 4.6981496167063174e-05,
+      "loss": 1.371,
+      "step": 11429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5644878149032593,
+      "learning_rate": 4.6981231826592655e-05,
+      "loss": 1.4656,
+      "step": 11430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7958176136016846,
+      "learning_rate": 4.698096748612213e-05,
+      "loss": 1.2823,
+      "step": 11431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.447354793548584,
+      "learning_rate": 4.69807031456516e-05,
+      "loss": 1.4075,
+      "step": 11432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.293393850326538,
+      "learning_rate": 4.6980438805181075e-05,
+      "loss": 1.4533,
+      "step": 11433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.995861530303955,
+      "learning_rate": 4.698017446471055e-05,
+      "loss": 1.4345,
+      "step": 11434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6268030405044556,
+      "learning_rate": 4.697991012424002e-05,
+      "loss": 1.5997,
+      "step": 11435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.37979793548584,
+      "learning_rate": 4.6979645783769496e-05,
+      "loss": 1.336,
+      "step": 11436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7074919939041138,
+      "learning_rate": 4.697938144329897e-05,
+      "loss": 2.6163,
+      "step": 11437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.48136305809021,
+      "learning_rate": 4.697911710282844e-05,
+      "loss": 1.152,
+      "step": 11438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.623479127883911,
+      "learning_rate": 4.697885276235792e-05,
+      "loss": 1.5955,
+      "step": 11439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9557684659957886,
+      "learning_rate": 4.697858842188739e-05,
+      "loss": 1.5288,
+      "step": 11440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.297359585762024,
+      "learning_rate": 4.697832408141687e-05,
+      "loss": 1.3572,
+      "step": 11441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5932741165161133,
+      "learning_rate": 4.697805974094634e-05,
+      "loss": 1.3532,
+      "step": 11442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8255808353424072,
+      "learning_rate": 4.697779540047582e-05,
+      "loss": 1.3926,
+      "step": 11443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8559902906417847,
+      "learning_rate": 4.6977531060005285e-05,
+      "loss": 1.4411,
+      "step": 11444
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2091987133026123,
+      "learning_rate": 4.6977266719534765e-05,
+      "loss": 1.2727,
+      "step": 11445
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5811678171157837,
+      "learning_rate": 4.697700237906424e-05,
+      "loss": 1.3695,
+      "step": 11446
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8652063608169556,
+      "learning_rate": 4.697673803859371e-05,
+      "loss": 2.5809,
+      "step": 11447
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4727779626846313,
+      "learning_rate": 4.6976473698123186e-05,
+      "loss": 1.4174,
+      "step": 11448
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0505030155181885,
+      "learning_rate": 4.697620935765266e-05,
+      "loss": 1.5199,
+      "step": 11449
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5112698078155518,
+      "learning_rate": 4.697594501718213e-05,
+      "loss": 1.3409,
+      "step": 11450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.70750892162323,
+      "learning_rate": 4.6975680676711606e-05,
+      "loss": 1.529,
+      "step": 11451
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.085390329360962,
+      "learning_rate": 4.697541633624108e-05,
+      "loss": 1.6332,
+      "step": 11452
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.589509129524231,
+      "learning_rate": 4.6975151995770554e-05,
+      "loss": 1.9543,
+      "step": 11453
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9589706659317017,
+      "learning_rate": 4.697488765530003e-05,
+      "loss": 1.4358,
+      "step": 11454
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9167014360427856,
+      "learning_rate": 4.69746233148295e-05,
+      "loss": 1.4278,
+      "step": 11455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.704898476600647,
+      "learning_rate": 4.697435897435898e-05,
+      "loss": 1.5094,
+      "step": 11456
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5285561084747314,
+      "learning_rate": 4.697409463388845e-05,
+      "loss": 1.4088,
+      "step": 11457
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2049448490142822,
+      "learning_rate": 4.697383029341793e-05,
+      "loss": 1.4665,
+      "step": 11458
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9956046342849731,
+      "learning_rate": 4.6973565952947395e-05,
+      "loss": 1.442,
+      "step": 11459
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8261364698410034,
+      "learning_rate": 4.6973301612476875e-05,
+      "loss": 1.4215,
+      "step": 11460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.002023935317993,
+      "learning_rate": 4.697303727200635e-05,
+      "loss": 1.2875,
+      "step": 11461
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9286468029022217,
+      "learning_rate": 4.6972772931535816e-05,
+      "loss": 1.4868,
+      "step": 11462
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6461149454116821,
+      "learning_rate": 4.6972508591065296e-05,
+      "loss": 1.2979,
+      "step": 11463
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5989456176757812,
+      "learning_rate": 4.697224425059476e-05,
+      "loss": 1.3281,
+      "step": 11464
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6515098810195923,
+      "learning_rate": 4.697197991012424e-05,
+      "loss": 1.4724,
+      "step": 11465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2644734382629395,
+      "learning_rate": 4.697171556965372e-05,
+      "loss": 1.4727,
+      "step": 11466
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9806878566741943,
+      "learning_rate": 4.697145122918319e-05,
+      "loss": 1.3276,
+      "step": 11467
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2347118854522705,
+      "learning_rate": 4.6971186888712664e-05,
+      "loss": 1.2234,
+      "step": 11468
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7092621326446533,
+      "learning_rate": 4.697092254824214e-05,
+      "loss": 1.3994,
+      "step": 11469
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.00203275680542,
+      "learning_rate": 4.697065820777161e-05,
+      "loss": 1.4217,
+      "step": 11470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5317310094833374,
+      "learning_rate": 4.6970393867301085e-05,
+      "loss": 2.0353,
+      "step": 11471
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6715935468673706,
+      "learning_rate": 4.697012952683056e-05,
+      "loss": 1.5399,
+      "step": 11472
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7129011154174805,
+      "learning_rate": 4.696986518636003e-05,
+      "loss": 1.3499,
+      "step": 11473
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8282663822174072,
+      "learning_rate": 4.6969600845889505e-05,
+      "loss": 1.5511,
+      "step": 11474
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6215312480926514,
+      "learning_rate": 4.696933650541898e-05,
+      "loss": 1.3349,
+      "step": 11475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3932092189788818,
+      "learning_rate": 4.696907216494846e-05,
+      "loss": 1.0918,
+      "step": 11476
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8444666862487793,
+      "learning_rate": 4.6968807824477926e-05,
+      "loss": 1.2364,
+      "step": 11477
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.007934808731079,
+      "learning_rate": 4.6968543484007406e-05,
+      "loss": 1.5058,
+      "step": 11478
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.83219575881958,
+      "learning_rate": 4.696827914353687e-05,
+      "loss": 1.4333,
+      "step": 11479
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4362815618515015,
+      "learning_rate": 4.6968014803066354e-05,
+      "loss": 1.3531,
+      "step": 11480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6019628047943115,
+      "learning_rate": 4.696775046259583e-05,
+      "loss": 1.8586,
+      "step": 11481
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0214173793792725,
+      "learning_rate": 4.69674861221253e-05,
+      "loss": 1.4022,
+      "step": 11482
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4280446767807007,
+      "learning_rate": 4.6967221781654774e-05,
+      "loss": 1.317,
+      "step": 11483
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9942477941513062,
+      "learning_rate": 4.696695744118425e-05,
+      "loss": 1.4002,
+      "step": 11484
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5946269035339355,
+      "learning_rate": 4.696669310071372e-05,
+      "loss": 1.6625,
+      "step": 11485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.819611668586731,
+      "learning_rate": 4.6966428760243195e-05,
+      "loss": 1.4003,
+      "step": 11486
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9081658124923706,
+      "learning_rate": 4.696616441977267e-05,
+      "loss": 1.4802,
+      "step": 11487
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2997210025787354,
+      "learning_rate": 4.696590007930214e-05,
+      "loss": 1.136,
+      "step": 11488
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1243391036987305,
+      "learning_rate": 4.6965635738831616e-05,
+      "loss": 1.5092,
+      "step": 11489
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8770831823349,
+      "learning_rate": 4.696537139836109e-05,
+      "loss": 1.3239,
+      "step": 11490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4672905206680298,
+      "learning_rate": 4.696510705789057e-05,
+      "loss": 1.2087,
+      "step": 11491
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9720261096954346,
+      "learning_rate": 4.6964842717420036e-05,
+      "loss": 1.4264,
+      "step": 11492
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4155092239379883,
+      "learning_rate": 4.696457837694952e-05,
+      "loss": 1.395,
+      "step": 11493
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8197498321533203,
+      "learning_rate": 4.6964314036478984e-05,
+      "loss": 1.4766,
+      "step": 11494
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8497215509414673,
+      "learning_rate": 4.6964049696008464e-05,
+      "loss": 2.0256,
+      "step": 11495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.120371103286743,
+      "learning_rate": 4.696378535553794e-05,
+      "loss": 1.4306,
+      "step": 11496
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9400379657745361,
+      "learning_rate": 4.696352101506741e-05,
+      "loss": 1.3461,
+      "step": 11497
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.008342742919922,
+      "learning_rate": 4.6963256674596885e-05,
+      "loss": 1.4539,
+      "step": 11498
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2260169982910156,
+      "learning_rate": 4.696299233412635e-05,
+      "loss": 1.5279,
+      "step": 11499
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.729105830192566,
+      "learning_rate": 4.696272799365583e-05,
+      "loss": 1.3268,
+      "step": 11500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4574596881866455,
+      "learning_rate": 4.6962463653185305e-05,
+      "loss": 1.5223,
+      "step": 11501
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5058581829071045,
+      "learning_rate": 4.696219931271478e-05,
+      "loss": 1.4964,
+      "step": 11502
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7231935262680054,
+      "learning_rate": 4.696193497224425e-05,
+      "loss": 1.3328,
+      "step": 11503
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3792612552642822,
+      "learning_rate": 4.6961670631773726e-05,
+      "loss": 1.5467,
+      "step": 11504
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7808412313461304,
+      "learning_rate": 4.69614062913032e-05,
+      "loss": 1.2389,
+      "step": 11505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9244352579116821,
+      "learning_rate": 4.696114195083268e-05,
+      "loss": 1.4536,
+      "step": 11506
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7028605937957764,
+      "learning_rate": 4.696087761036215e-05,
+      "loss": 1.4189,
+      "step": 11507
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9628541469573975,
+      "learning_rate": 4.696061326989162e-05,
+      "loss": 1.9571,
+      "step": 11508
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4347032308578491,
+      "learning_rate": 4.6960348929421094e-05,
+      "loss": 1.3666,
+      "step": 11509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0543668270111084,
+      "learning_rate": 4.696008458895057e-05,
+      "loss": 1.42,
+      "step": 11510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.836374044418335,
+      "learning_rate": 4.695982024848005e-05,
+      "loss": 1.3331,
+      "step": 11511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9314846992492676,
+      "learning_rate": 4.6959555908009515e-05,
+      "loss": 1.4791,
+      "step": 11512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9292876720428467,
+      "learning_rate": 4.6959291567538995e-05,
+      "loss": 1.4092,
+      "step": 11513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7812628746032715,
+      "learning_rate": 4.695902722706846e-05,
+      "loss": 1.3435,
+      "step": 11514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6699146032333374,
+      "learning_rate": 4.695876288659794e-05,
+      "loss": 1.3331,
+      "step": 11515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5003855228424072,
+      "learning_rate": 4.6958498546127416e-05,
+      "loss": 1.2247,
+      "step": 11516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9482460021972656,
+      "learning_rate": 4.695823420565689e-05,
+      "loss": 1.6446,
+      "step": 11517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5476830005645752,
+      "learning_rate": 4.695796986518636e-05,
+      "loss": 1.2888,
+      "step": 11518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9233359098434448,
+      "learning_rate": 4.6957705524715836e-05,
+      "loss": 1.5395,
+      "step": 11519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.618739128112793,
+      "learning_rate": 4.695744118424531e-05,
+      "loss": 1.3826,
+      "step": 11520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2890970706939697,
+      "learning_rate": 4.6957176843774783e-05,
+      "loss": 1.5901,
+      "step": 11521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6118417978286743,
+      "learning_rate": 4.695691250330426e-05,
+      "loss": 1.3282,
+      "step": 11522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.530276894569397,
+      "learning_rate": 4.695664816283373e-05,
+      "loss": 1.2827,
+      "step": 11523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.925882339477539,
+      "learning_rate": 4.6956383822363204e-05,
+      "loss": 1.3532,
+      "step": 11524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5897036790847778,
+      "learning_rate": 4.695611948189268e-05,
+      "loss": 1.1904,
+      "step": 11525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.640950083732605,
+      "learning_rate": 4.695585514142216e-05,
+      "loss": 1.3467,
+      "step": 11526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2048604488372803,
+      "learning_rate": 4.6955590800951625e-05,
+      "loss": 1.452,
+      "step": 11527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7499616146087646,
+      "learning_rate": 4.6955326460481105e-05,
+      "loss": 1.2958,
+      "step": 11528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3410053253173828,
+      "learning_rate": 4.695506212001057e-05,
+      "loss": 1.3589,
+      "step": 11529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8911018371582031,
+      "learning_rate": 4.695479777954005e-05,
+      "loss": 1.4265,
+      "step": 11530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9250004291534424,
+      "learning_rate": 4.6954533439069526e-05,
+      "loss": 1.6757,
+      "step": 11531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4594136476516724,
+      "learning_rate": 4.6954269098599e-05,
+      "loss": 1.3226,
+      "step": 11532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0623412132263184,
+      "learning_rate": 4.695400475812847e-05,
+      "loss": 1.3822,
+      "step": 11533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.514304757118225,
+      "learning_rate": 4.695374041765794e-05,
+      "loss": 1.2253,
+      "step": 11534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5857094526290894,
+      "learning_rate": 4.695347607718742e-05,
+      "loss": 1.7413,
+      "step": 11535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7140848636627197,
+      "learning_rate": 4.6953211736716894e-05,
+      "loss": 1.3805,
+      "step": 11536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7632888555526733,
+      "learning_rate": 4.695294739624637e-05,
+      "loss": 1.5452,
+      "step": 11537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5939068794250488,
+      "learning_rate": 4.695268305577584e-05,
+      "loss": 1.3266,
+      "step": 11538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5885835886001587,
+      "learning_rate": 4.6952418715305315e-05,
+      "loss": 1.4022,
+      "step": 11539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5689297914505005,
+      "learning_rate": 4.695215437483479e-05,
+      "loss": 1.4225,
+      "step": 11540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6519255638122559,
+      "learning_rate": 4.695189003436427e-05,
+      "loss": 1.4122,
+      "step": 11541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.864654779434204,
+      "learning_rate": 4.6951625693893735e-05,
+      "loss": 1.4295,
+      "step": 11542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5725023746490479,
+      "learning_rate": 4.695136135342321e-05,
+      "loss": 1.2732,
+      "step": 11543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.502059817314148,
+      "learning_rate": 4.695109701295268e-05,
+      "loss": 1.464,
+      "step": 11544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6778416633605957,
+      "learning_rate": 4.6950832672482156e-05,
+      "loss": 1.4501,
+      "step": 11545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2596755027770996,
+      "learning_rate": 4.6950568332011636e-05,
+      "loss": 1.4784,
+      "step": 11546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.837555170059204,
+      "learning_rate": 4.69503039915411e-05,
+      "loss": 1.3435,
+      "step": 11547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8078925609588623,
+      "learning_rate": 4.6950039651070583e-05,
+      "loss": 1.2683,
+      "step": 11548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7580236196517944,
+      "learning_rate": 4.694977531060005e-05,
+      "loss": 1.2713,
+      "step": 11549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1442363262176514,
+      "learning_rate": 4.694951097012953e-05,
+      "loss": 2.7938,
+      "step": 11550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6925557851791382,
+      "learning_rate": 4.6949246629659004e-05,
+      "loss": 1.3608,
+      "step": 11551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.017533540725708,
+      "learning_rate": 4.694898228918848e-05,
+      "loss": 1.5107,
+      "step": 11552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4658116102218628,
+      "learning_rate": 4.694871794871795e-05,
+      "loss": 1.361,
+      "step": 11553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5422711372375488,
+      "learning_rate": 4.6948453608247425e-05,
+      "loss": 1.3547,
+      "step": 11554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.793718695640564,
+      "learning_rate": 4.69481892677769e-05,
+      "loss": 2.458,
+      "step": 11555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.299565315246582,
+      "learning_rate": 4.694792492730637e-05,
+      "loss": 1.6135,
+      "step": 11556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.221348524093628,
+      "learning_rate": 4.6947660586835846e-05,
+      "loss": 1.4822,
+      "step": 11557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0425357818603516,
+      "learning_rate": 4.694739624636532e-05,
+      "loss": 1.5438,
+      "step": 11558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5407607555389404,
+      "learning_rate": 4.694713190589479e-05,
+      "loss": 1.51,
+      "step": 11559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.392013430595398,
+      "learning_rate": 4.6946867565424266e-05,
+      "loss": 1.4074,
+      "step": 11560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.606566309928894,
+      "learning_rate": 4.694660322495375e-05,
+      "loss": 1.2344,
+      "step": 11561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4461796283721924,
+      "learning_rate": 4.6946338884483213e-05,
+      "loss": 1.5443,
+      "step": 11562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7067302465438843,
+      "learning_rate": 4.6946074544012694e-05,
+      "loss": 1.3571,
+      "step": 11563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4990395307540894,
+      "learning_rate": 4.694581020354216e-05,
+      "loss": 1.399,
+      "step": 11564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7227821350097656,
+      "learning_rate": 4.694554586307164e-05,
+      "loss": 1.4327,
+      "step": 11565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.975823163986206,
+      "learning_rate": 4.6945281522601115e-05,
+      "loss": 1.4722,
+      "step": 11566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.881903886795044,
+      "learning_rate": 4.694501718213059e-05,
+      "loss": 1.5096,
+      "step": 11567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.536894679069519,
+      "learning_rate": 4.694475284166006e-05,
+      "loss": 1.3332,
+      "step": 11568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6025731563568115,
+      "learning_rate": 4.6944488501189535e-05,
+      "loss": 1.4412,
+      "step": 11569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4940080642700195,
+      "learning_rate": 4.694422416071901e-05,
+      "loss": 1.2637,
+      "step": 11570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7382060289382935,
+      "learning_rate": 4.694395982024848e-05,
+      "loss": 1.4608,
+      "step": 11571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9321519136428833,
+      "learning_rate": 4.6943695479777956e-05,
+      "loss": 1.4806,
+      "step": 11572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6538299322128296,
+      "learning_rate": 4.694343113930743e-05,
+      "loss": 1.4201,
+      "step": 11573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3286705017089844,
+      "learning_rate": 4.69431667988369e-05,
+      "loss": 1.4926,
+      "step": 11574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3875529766082764,
+      "learning_rate": 4.694290245836638e-05,
+      "loss": 1.273,
+      "step": 11575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5282254219055176,
+      "learning_rate": 4.694263811789586e-05,
+      "loss": 1.4074,
+      "step": 11576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.790358543395996,
+      "learning_rate": 4.6942373777425324e-05,
+      "loss": 1.3954,
+      "step": 11577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5184471607208252,
+      "learning_rate": 4.6942109436954804e-05,
+      "loss": 1.293,
+      "step": 11578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7086374759674072,
+      "learning_rate": 4.694184509648427e-05,
+      "loss": 1.4447,
+      "step": 11579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8555017709732056,
+      "learning_rate": 4.6941580756013745e-05,
+      "loss": 1.4046,
+      "step": 11580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4288073778152466,
+      "learning_rate": 4.6941316415543225e-05,
+      "loss": 1.197,
+      "step": 11581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0484914779663086,
+      "learning_rate": 4.694105207507269e-05,
+      "loss": 1.4615,
+      "step": 11582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8446357250213623,
+      "learning_rate": 4.694078773460217e-05,
+      "loss": 1.6414,
+      "step": 11583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.745840311050415,
+      "learning_rate": 4.694052339413164e-05,
+      "loss": 1.5281,
+      "step": 11584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.437429904937744,
+      "learning_rate": 4.694025905366112e-05,
+      "loss": 1.4682,
+      "step": 11585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8160088062286377,
+      "learning_rate": 4.693999471319059e-05,
+      "loss": 1.2842,
+      "step": 11586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.712257981300354,
+      "learning_rate": 4.6939730372720066e-05,
+      "loss": 1.449,
+      "step": 11587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2396152019500732,
+      "learning_rate": 4.693946603224954e-05,
+      "loss": 1.3231,
+      "step": 11588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9234551191329956,
+      "learning_rate": 4.6939201691779013e-05,
+      "loss": 1.4398,
+      "step": 11589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9418561458587646,
+      "learning_rate": 4.693893735130849e-05,
+      "loss": 1.4278,
+      "step": 11590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3613314628601074,
+      "learning_rate": 4.693867301083796e-05,
+      "loss": 1.3174,
+      "step": 11591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8629313707351685,
+      "learning_rate": 4.6938408670367434e-05,
+      "loss": 1.4094,
+      "step": 11592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9932061433792114,
+      "learning_rate": 4.693814432989691e-05,
+      "loss": 1.44,
+      "step": 11593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.626163125038147,
+      "learning_rate": 4.693787998942638e-05,
+      "loss": 1.3975,
+      "step": 11594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9012787342071533,
+      "learning_rate": 4.6937615648955855e-05,
+      "loss": 1.4054,
+      "step": 11595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9245208501815796,
+      "learning_rate": 4.6937351308485335e-05,
+      "loss": 1.3403,
+      "step": 11596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7024873495101929,
+      "learning_rate": 4.69370869680148e-05,
+      "loss": 1.337,
+      "step": 11597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4634737968444824,
+      "learning_rate": 4.693682262754428e-05,
+      "loss": 1.5343,
+      "step": 11598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8252506256103516,
+      "learning_rate": 4.693655828707375e-05,
+      "loss": 1.3509,
+      "step": 11599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.539733648300171,
+      "learning_rate": 4.693629394660323e-05,
+      "loss": 1.3167,
+      "step": 11600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0241339206695557,
+      "learning_rate": 4.69360296061327e-05,
+      "loss": 1.4379,
+      "step": 11601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.567167043685913,
+      "learning_rate": 4.693576526566218e-05,
+      "loss": 1.3303,
+      "step": 11602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6347380876541138,
+      "learning_rate": 4.693550092519165e-05,
+      "loss": 1.3757,
+      "step": 11603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5776336193084717,
+      "learning_rate": 4.6935236584721124e-05,
+      "loss": 1.2384,
+      "step": 11604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.50384783744812,
+      "learning_rate": 4.69349722442506e-05,
+      "loss": 1.3985,
+      "step": 11605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6618168354034424,
+      "learning_rate": 4.693470790378007e-05,
+      "loss": 1.35,
+      "step": 11606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.609166145324707,
+      "learning_rate": 4.6934443563309544e-05,
+      "loss": 1.4029,
+      "step": 11607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.749674916267395,
+      "learning_rate": 4.693417922283902e-05,
+      "loss": 1.4288,
+      "step": 11608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7121646404266357,
+      "learning_rate": 4.693391488236849e-05,
+      "loss": 1.3458,
+      "step": 11609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3286290168762207,
+      "learning_rate": 4.6933650541897965e-05,
+      "loss": 1.4874,
+      "step": 11610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.534224033355713,
+      "learning_rate": 4.6933386201427446e-05,
+      "loss": 1.3368,
+      "step": 11611
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6077955961227417,
+      "learning_rate": 4.693312186095691e-05,
+      "loss": 1.2835,
+      "step": 11612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5959707498550415,
+      "learning_rate": 4.693285752048639e-05,
+      "loss": 1.2476,
+      "step": 11613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7874125242233276,
+      "learning_rate": 4.693259318001586e-05,
+      "loss": 1.4113,
+      "step": 11614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.660762906074524,
+      "learning_rate": 4.693232883954534e-05,
+      "loss": 1.3875,
+      "step": 11615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8139889240264893,
+      "learning_rate": 4.693206449907481e-05,
+      "loss": 1.4678,
+      "step": 11616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9613479375839233,
+      "learning_rate": 4.693180015860428e-05,
+      "loss": 1.3787,
+      "step": 11617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3580102920532227,
+      "learning_rate": 4.693153581813376e-05,
+      "loss": 1.4672,
+      "step": 11618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6267989873886108,
+      "learning_rate": 4.693127147766323e-05,
+      "loss": 1.5273,
+      "step": 11619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1703414916992188,
+      "learning_rate": 4.693100713719271e-05,
+      "loss": 1.3766,
+      "step": 11620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9497264623641968,
+      "learning_rate": 4.693074279672218e-05,
+      "loss": 1.4068,
+      "step": 11621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9895650148391724,
+      "learning_rate": 4.6930478456251655e-05,
+      "loss": 1.3691,
+      "step": 11622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7729320526123047,
+      "learning_rate": 4.693021411578113e-05,
+      "loss": 1.3743,
+      "step": 11623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4501540660858154,
+      "learning_rate": 4.69299497753106e-05,
+      "loss": 1.2937,
+      "step": 11624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5813791751861572,
+      "learning_rate": 4.6929685434840076e-05,
+      "loss": 1.3638,
+      "step": 11625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.897799015045166,
+      "learning_rate": 4.692942109436955e-05,
+      "loss": 1.4689,
+      "step": 11626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6568264961242676,
+      "learning_rate": 4.692915675389902e-05,
+      "loss": 1.3812,
+      "step": 11627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7190568447113037,
+      "learning_rate": 4.6928892413428496e-05,
+      "loss": 1.4509,
+      "step": 11628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0567240715026855,
+      "learning_rate": 4.692862807295797e-05,
+      "loss": 1.3328,
+      "step": 11629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453685760498047,
+      "learning_rate": 4.692836373248744e-05,
+      "loss": 1.5745,
+      "step": 11630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.448885917663574,
+      "learning_rate": 4.6928099392016924e-05,
+      "loss": 1.5507,
+      "step": 11631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6725597381591797,
+      "learning_rate": 4.692783505154639e-05,
+      "loss": 1.4614,
+      "step": 11632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8089679479599,
+      "learning_rate": 4.692757071107587e-05,
+      "loss": 1.536,
+      "step": 11633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7098033428192139,
+      "learning_rate": 4.692730637060534e-05,
+      "loss": 1.3821,
+      "step": 11634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5676062107086182,
+      "learning_rate": 4.692704203013482e-05,
+      "loss": 1.1923,
+      "step": 11635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7166815996170044,
+      "learning_rate": 4.692677768966429e-05,
+      "loss": 1.4929,
+      "step": 11636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6863343715667725,
+      "learning_rate": 4.6926513349193765e-05,
+      "loss": 1.3501,
+      "step": 11637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6050488948822021,
+      "learning_rate": 4.692624900872324e-05,
+      "loss": 1.3295,
+      "step": 11638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7914711236953735,
+      "learning_rate": 4.692598466825271e-05,
+      "loss": 1.1769,
+      "step": 11639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.012338876724243,
+      "learning_rate": 4.6925720327782186e-05,
+      "loss": 1.3652,
+      "step": 11640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.349283218383789,
+      "learning_rate": 4.692545598731166e-05,
+      "loss": 1.4272,
+      "step": 11641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1255440711975098,
+      "learning_rate": 4.692519164684113e-05,
+      "loss": 1.4647,
+      "step": 11642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.334831476211548,
+      "learning_rate": 4.6924927306370607e-05,
+      "loss": 1.4664,
+      "step": 11643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.417060136795044,
+      "learning_rate": 4.692466296590008e-05,
+      "loss": 1.1427,
+      "step": 11644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.628499984741211,
+      "learning_rate": 4.6924398625429554e-05,
+      "loss": 1.2882,
+      "step": 11645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3964283466339111,
+      "learning_rate": 4.6924134284959034e-05,
+      "loss": 1.2614,
+      "step": 11646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2964444160461426,
+      "learning_rate": 4.69238699444885e-05,
+      "loss": 1.3577,
+      "step": 11647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8732532262802124,
+      "learning_rate": 4.692360560401798e-05,
+      "loss": 1.505,
+      "step": 11648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5581218004226685,
+      "learning_rate": 4.692334126354745e-05,
+      "loss": 1.3277,
+      "step": 11649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1024134159088135,
+      "learning_rate": 4.692307692307693e-05,
+      "loss": 1.2371,
+      "step": 11650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03096079826355,
+      "learning_rate": 4.69228125826064e-05,
+      "loss": 1.4889,
+      "step": 11651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.053443431854248,
+      "learning_rate": 4.692254824213587e-05,
+      "loss": 1.3613,
+      "step": 11652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1770083904266357,
+      "learning_rate": 4.692228390166535e-05,
+      "loss": 1.4247,
+      "step": 11653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.471248984336853,
+      "learning_rate": 4.6922019561194816e-05,
+      "loss": 1.3456,
+      "step": 11654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.686184287071228,
+      "learning_rate": 4.6921755220724296e-05,
+      "loss": 1.4716,
+      "step": 11655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6265296936035156,
+      "learning_rate": 4.692149088025377e-05,
+      "loss": 1.293,
+      "step": 11656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.893489122390747,
+      "learning_rate": 4.692122653978324e-05,
+      "loss": 1.1671,
+      "step": 11657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9047226905822754,
+      "learning_rate": 4.692096219931272e-05,
+      "loss": 1.4664,
+      "step": 11658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7178140878677368,
+      "learning_rate": 4.692069785884219e-05,
+      "loss": 1.4547,
+      "step": 11659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9641746282577515,
+      "learning_rate": 4.6920433518371664e-05,
+      "loss": 1.3712,
+      "step": 11660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.499127149581909,
+      "learning_rate": 4.692016917790114e-05,
+      "loss": 1.5193,
+      "step": 11661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8220101594924927,
+      "learning_rate": 4.691990483743061e-05,
+      "loss": 1.2516,
+      "step": 11662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.10898494720459,
+      "learning_rate": 4.6919640496960085e-05,
+      "loss": 1.4642,
+      "step": 11663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7430927753448486,
+      "learning_rate": 4.691937615648956e-05,
+      "loss": 1.4242,
+      "step": 11664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.891067624092102,
+      "learning_rate": 4.691911181601903e-05,
+      "loss": 1.4677,
+      "step": 11665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4679814577102661,
+      "learning_rate": 4.691884747554851e-05,
+      "loss": 1.4366,
+      "step": 11666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7033028602600098,
+      "learning_rate": 4.691858313507798e-05,
+      "loss": 1.4425,
+      "step": 11667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8963085412979126,
+      "learning_rate": 4.691831879460746e-05,
+      "loss": 1.393,
+      "step": 11668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7396726608276367,
+      "learning_rate": 4.6918054454136926e-05,
+      "loss": 1.3865,
+      "step": 11669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.471248745918274,
+      "learning_rate": 4.6917790113666407e-05,
+      "loss": 2.0742,
+      "step": 11670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8426496982574463,
+      "learning_rate": 4.691752577319588e-05,
+      "loss": 1.4445,
+      "step": 11671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6662427186965942,
+      "learning_rate": 4.6917261432725354e-05,
+      "loss": 1.2077,
+      "step": 11672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.935653805732727,
+      "learning_rate": 4.691699709225483e-05,
+      "loss": 1.3074,
+      "step": 11673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6252920627593994,
+      "learning_rate": 4.69167327517843e-05,
+      "loss": 1.607,
+      "step": 11674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.954100251197815,
+      "learning_rate": 4.6916468411313774e-05,
+      "loss": 1.3562,
+      "step": 11675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.668614149093628,
+      "learning_rate": 4.691620407084325e-05,
+      "loss": 1.3157,
+      "step": 11676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9129183292388916,
+      "learning_rate": 4.691593973037272e-05,
+      "loss": 1.4605,
+      "step": 11677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.918936014175415,
+      "learning_rate": 4.6915675389902195e-05,
+      "loss": 1.3643,
+      "step": 11678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9469060897827148,
+      "learning_rate": 4.691541104943167e-05,
+      "loss": 1.3747,
+      "step": 11679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0813839435577393,
+      "learning_rate": 4.691514670896114e-05,
+      "loss": 1.3426,
+      "step": 11680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.741990566253662,
+      "learning_rate": 4.691488236849062e-05,
+      "loss": 1.3298,
+      "step": 11681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.150705575942993,
+      "learning_rate": 4.691461802802009e-05,
+      "loss": 1.3061,
+      "step": 11682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.7948994636535645,
+      "learning_rate": 4.691435368754957e-05,
+      "loss": 1.5105,
+      "step": 11683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7933573722839355,
+      "learning_rate": 4.6914089347079037e-05,
+      "loss": 1.7255,
+      "step": 11684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4911032915115356,
+      "learning_rate": 4.691382500660852e-05,
+      "loss": 1.3277,
+      "step": 11685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6566275358200073,
+      "learning_rate": 4.691356066613799e-05,
+      "loss": 1.1976,
+      "step": 11686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9003069400787354,
+      "learning_rate": 4.6913296325667464e-05,
+      "loss": 1.57,
+      "step": 11687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6568022966384888,
+      "learning_rate": 4.691303198519694e-05,
+      "loss": 1.3241,
+      "step": 11688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3695383071899414,
+      "learning_rate": 4.6912767644726404e-05,
+      "loss": 1.5382,
+      "step": 11689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4089981317520142,
+      "learning_rate": 4.6912503304255885e-05,
+      "loss": 1.6909,
+      "step": 11690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.013536214828491,
+      "learning_rate": 4.691223896378536e-05,
+      "loss": 1.9355,
+      "step": 11691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.239962100982666,
+      "learning_rate": 4.691197462331483e-05,
+      "loss": 1.5647,
+      "step": 11692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9295766353607178,
+      "learning_rate": 4.6911710282844305e-05,
+      "loss": 1.5267,
+      "step": 11693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8160977363586426,
+      "learning_rate": 4.691144594237378e-05,
+      "loss": 1.499,
+      "step": 11694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8396151065826416,
+      "learning_rate": 4.691118160190325e-05,
+      "loss": 1.3073,
+      "step": 11695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8260407447814941,
+      "learning_rate": 4.691091726143273e-05,
+      "loss": 1.4241,
+      "step": 11696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4925520420074463,
+      "learning_rate": 4.69106529209622e-05,
+      "loss": 1.3857,
+      "step": 11697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7283973693847656,
+      "learning_rate": 4.691038858049167e-05,
+      "loss": 1.4127,
+      "step": 11698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9689651727676392,
+      "learning_rate": 4.691012424002115e-05,
+      "loss": 1.4444,
+      "step": 11699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.687425136566162,
+      "learning_rate": 4.690985989955062e-05,
+      "loss": 1.3818,
+      "step": 11700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.268343448638916,
+      "learning_rate": 4.69095955590801e-05,
+      "loss": 1.4443,
+      "step": 11701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8710254430770874,
+      "learning_rate": 4.690933121860957e-05,
+      "loss": 1.423,
+      "step": 11702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.703936219215393,
+      "learning_rate": 4.690906687813905e-05,
+      "loss": 1.3204,
+      "step": 11703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.039677143096924,
+      "learning_rate": 4.6908802537668515e-05,
+      "loss": 1.6065,
+      "step": 11704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6540296077728271,
+      "learning_rate": 4.6908538197197995e-05,
+      "loss": 1.4344,
+      "step": 11705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1534066200256348,
+      "learning_rate": 4.690827385672747e-05,
+      "loss": 1.2364,
+      "step": 11706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.184147596359253,
+      "learning_rate": 4.690800951625694e-05,
+      "loss": 1.5255,
+      "step": 11707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7120853662490845,
+      "learning_rate": 4.6907745175786416e-05,
+      "loss": 1.3966,
+      "step": 11708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8341299295425415,
+      "learning_rate": 4.690748083531589e-05,
+      "loss": 1.4899,
+      "step": 11709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.023836135864258,
+      "learning_rate": 4.690721649484536e-05,
+      "loss": 1.4744,
+      "step": 11710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0415000915527344,
+      "learning_rate": 4.6906952154374837e-05,
+      "loss": 1.3277,
+      "step": 11711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.921911358833313,
+      "learning_rate": 4.690668781390431e-05,
+      "loss": 1.4371,
+      "step": 11712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6559219360351562,
+      "learning_rate": 4.6906423473433784e-05,
+      "loss": 1.3433,
+      "step": 11713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0194785594940186,
+      "learning_rate": 4.690615913296326e-05,
+      "loss": 1.4206,
+      "step": 11714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.401432991027832,
+      "learning_rate": 4.690589479249273e-05,
+      "loss": 1.2,
+      "step": 11715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.130389451980591,
+      "learning_rate": 4.690563045202221e-05,
+      "loss": 1.5733,
+      "step": 11716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.294245958328247,
+      "learning_rate": 4.690536611155168e-05,
+      "loss": 1.3822,
+      "step": 11717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7063324451446533,
+      "learning_rate": 4.690510177108116e-05,
+      "loss": 1.3676,
+      "step": 11718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.696897029876709,
+      "learning_rate": 4.6904837430610625e-05,
+      "loss": 1.4044,
+      "step": 11719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.451676368713379,
+      "learning_rate": 4.6904573090140105e-05,
+      "loss": 1.468,
+      "step": 11720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6925593614578247,
+      "learning_rate": 4.690430874966958e-05,
+      "loss": 1.4674,
+      "step": 11721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.180718183517456,
+      "learning_rate": 4.690404440919905e-05,
+      "loss": 1.6057,
+      "step": 11722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5384974479675293,
+      "learning_rate": 4.6903780068728526e-05,
+      "loss": 1.416,
+      "step": 11723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5125486850738525,
+      "learning_rate": 4.690351572825799e-05,
+      "loss": 1.2459,
+      "step": 11724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8421343564987183,
+      "learning_rate": 4.690325138778747e-05,
+      "loss": 1.379,
+      "step": 11725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0755326747894287,
+      "learning_rate": 4.690298704731695e-05,
+      "loss": 1.4272,
+      "step": 11726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.322953462600708,
+      "learning_rate": 4.690272270684642e-05,
+      "loss": 1.4709,
+      "step": 11727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.878684639930725,
+      "learning_rate": 4.6902458366375894e-05,
+      "loss": 1.4527,
+      "step": 11728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4377245903015137,
+      "learning_rate": 4.690219402590537e-05,
+      "loss": 1.371,
+      "step": 11729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.812878131866455,
+      "learning_rate": 4.690192968543484e-05,
+      "loss": 1.4885,
+      "step": 11730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9835031032562256,
+      "learning_rate": 4.690166534496432e-05,
+      "loss": 1.3472,
+      "step": 11731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.374194622039795,
+      "learning_rate": 4.690140100449379e-05,
+      "loss": 1.2087,
+      "step": 11732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0435359477996826,
+      "learning_rate": 4.690113666402326e-05,
+      "loss": 1.4941,
+      "step": 11733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8187861442565918,
+      "learning_rate": 4.6900872323552735e-05,
+      "loss": 1.4327,
+      "step": 11734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.26609206199646,
+      "learning_rate": 4.690060798308221e-05,
+      "loss": 1.6149,
+      "step": 11735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.198056221008301,
+      "learning_rate": 4.690034364261169e-05,
+      "loss": 1.5142,
+      "step": 11736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8961478471755981,
+      "learning_rate": 4.6900079302141156e-05,
+      "loss": 1.2841,
+      "step": 11737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.522771954536438,
+      "learning_rate": 4.6899814961670636e-05,
+      "loss": 1.1368,
+      "step": 11738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3844209909439087,
+      "learning_rate": 4.68995506212001e-05,
+      "loss": 1.2719,
+      "step": 11739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7884855270385742,
+      "learning_rate": 4.6899286280729584e-05,
+      "loss": 1.4207,
+      "step": 11740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7149120569229126,
+      "learning_rate": 4.689902194025906e-05,
+      "loss": 1.371,
+      "step": 11741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6846996545791626,
+      "learning_rate": 4.689875759978853e-05,
+      "loss": 1.2675,
+      "step": 11742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5128611326217651,
+      "learning_rate": 4.6898493259318004e-05,
+      "loss": 1.3101,
+      "step": 11743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.405944585800171,
+      "learning_rate": 4.689822891884748e-05,
+      "loss": 1.3066,
+      "step": 11744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8556522130966187,
+      "learning_rate": 4.689796457837695e-05,
+      "loss": 1.4397,
+      "step": 11745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5407352447509766,
+      "learning_rate": 4.6897700237906425e-05,
+      "loss": 1.4115,
+      "step": 11746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1381590366363525,
+      "learning_rate": 4.68974358974359e-05,
+      "loss": 1.3136,
+      "step": 11747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6094160079956055,
+      "learning_rate": 4.689717155696537e-05,
+      "loss": 1.3456,
+      "step": 11748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1037542819976807,
+      "learning_rate": 4.6896907216494846e-05,
+      "loss": 1.6991,
+      "step": 11749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4465513229370117,
+      "learning_rate": 4.689664287602432e-05,
+      "loss": 1.4456,
+      "step": 11750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.198232412338257,
+      "learning_rate": 4.68963785355538e-05,
+      "loss": 1.4406,
+      "step": 11751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.921389102935791,
+      "learning_rate": 4.6896114195083266e-05,
+      "loss": 1.3066,
+      "step": 11752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7765843868255615,
+      "learning_rate": 4.689584985461275e-05,
+      "loss": 1.5069,
+      "step": 11753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8593547344207764,
+      "learning_rate": 4.6895585514142214e-05,
+      "loss": 1.4932,
+      "step": 11754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1371548175811768,
+      "learning_rate": 4.6895321173671694e-05,
+      "loss": 1.3148,
+      "step": 11755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0462465286254883,
+      "learning_rate": 4.689505683320117e-05,
+      "loss": 1.4675,
+      "step": 11756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4689031839370728,
+      "learning_rate": 4.689479249273064e-05,
+      "loss": 1.272,
+      "step": 11757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.211078643798828,
+      "learning_rate": 4.6894528152260115e-05,
+      "loss": 1.5958,
+      "step": 11758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2285022735595703,
+      "learning_rate": 4.689426381178959e-05,
+      "loss": 1.4602,
+      "step": 11759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5425361394882202,
+      "learning_rate": 4.689399947131906e-05,
+      "loss": 1.4353,
+      "step": 11760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3668384552001953,
+      "learning_rate": 4.6893735130848535e-05,
+      "loss": 1.2046,
+      "step": 11761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2495484352111816,
+      "learning_rate": 4.689347079037801e-05,
+      "loss": 1.5665,
+      "step": 11762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4083480834960938,
+      "learning_rate": 4.689320644990748e-05,
+      "loss": 1.4753,
+      "step": 11763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2177271842956543,
+      "learning_rate": 4.6892942109436956e-05,
+      "loss": 1.4887,
+      "step": 11764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7769534587860107,
+      "learning_rate": 4.689267776896643e-05,
+      "loss": 1.5092,
+      "step": 11765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0374674797058105,
+      "learning_rate": 4.689241342849591e-05,
+      "loss": 1.4799,
+      "step": 11766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5835964679718018,
+      "learning_rate": 4.689214908802538e-05,
+      "loss": 2.6078,
+      "step": 11767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7926836013793945,
+      "learning_rate": 4.689188474755486e-05,
+      "loss": 1.3992,
+      "step": 11768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.661574363708496,
+      "learning_rate": 4.6891620407084324e-05,
+      "loss": 1.338,
+      "step": 11769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8644400835037231,
+      "learning_rate": 4.68913560666138e-05,
+      "loss": 1.3801,
+      "step": 11770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.449887752532959,
+      "learning_rate": 4.689109172614328e-05,
+      "loss": 1.5158,
+      "step": 11771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4571282863616943,
+      "learning_rate": 4.6890827385672745e-05,
+      "loss": 1.4685,
+      "step": 11772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4772063493728638,
+      "learning_rate": 4.6890563045202225e-05,
+      "loss": 1.3036,
+      "step": 11773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.273923873901367,
+      "learning_rate": 4.689029870473169e-05,
+      "loss": 1.7035,
+      "step": 11774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6714575290679932,
+      "learning_rate": 4.689003436426117e-05,
+      "loss": 1.3894,
+      "step": 11775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.722227692604065,
+      "learning_rate": 4.6889770023790646e-05,
+      "loss": 1.4267,
+      "step": 11776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9622403383255005,
+      "learning_rate": 4.688950568332012e-05,
+      "loss": 1.2725,
+      "step": 11777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0548770427703857,
+      "learning_rate": 4.688924134284959e-05,
+      "loss": 1.507,
+      "step": 11778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6897923946380615,
+      "learning_rate": 4.6888977002379066e-05,
+      "loss": 1.3056,
+      "step": 11779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1076953411102295,
+      "learning_rate": 4.688871266190854e-05,
+      "loss": 1.4092,
+      "step": 11780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9371975660324097,
+      "learning_rate": 4.6888448321438014e-05,
+      "loss": 1.4495,
+      "step": 11781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4723788499832153,
+      "learning_rate": 4.688818398096749e-05,
+      "loss": 1.4667,
+      "step": 11782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7227370738983154,
+      "learning_rate": 4.688791964049696e-05,
+      "loss": 2.3662,
+      "step": 11783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.033798933029175,
+      "learning_rate": 4.6887655300026434e-05,
+      "loss": 1.4633,
+      "step": 11784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.669054627418518,
+      "learning_rate": 4.688739095955591e-05,
+      "loss": 1.3224,
+      "step": 11785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6802079677581787,
+      "learning_rate": 4.688712661908539e-05,
+      "loss": 1.4032,
+      "step": 11786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.304321765899658,
+      "learning_rate": 4.6886862278614855e-05,
+      "loss": 1.5303,
+      "step": 11787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9964091777801514,
+      "learning_rate": 4.6886597938144335e-05,
+      "loss": 1.3209,
+      "step": 11788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8757359981536865,
+      "learning_rate": 4.68863335976738e-05,
+      "loss": 1.4674,
+      "step": 11789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.802573561668396,
+      "learning_rate": 4.688606925720328e-05,
+      "loss": 1.3615,
+      "step": 11790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8167829513549805,
+      "learning_rate": 4.6885804916732756e-05,
+      "loss": 1.3499,
+      "step": 11791
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8471299409866333,
+      "learning_rate": 4.688554057626223e-05,
+      "loss": 1.4463,
+      "step": 11792
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8594117164611816,
+      "learning_rate": 4.68852762357917e-05,
+      "loss": 1.4277,
+      "step": 11793
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8648897409439087,
+      "learning_rate": 4.688501189532118e-05,
+      "loss": 1.367,
+      "step": 11794
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7886059284210205,
+      "learning_rate": 4.688474755485065e-05,
+      "loss": 1.4381,
+      "step": 11795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7784500122070312,
+      "learning_rate": 4.6884483214380124e-05,
+      "loss": 1.3695,
+      "step": 11796
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.533725380897522,
+      "learning_rate": 4.68842188739096e-05,
+      "loss": 1.2296,
+      "step": 11797
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.284083127975464,
+      "learning_rate": 4.688395453343907e-05,
+      "loss": 1.6536,
+      "step": 11798
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7171120643615723,
+      "learning_rate": 4.6883690192968545e-05,
+      "loss": 1.5177,
+      "step": 11799
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0464115142822266,
+      "learning_rate": 4.688342585249802e-05,
+      "loss": 1.6104,
+      "step": 11800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0255746841430664,
+      "learning_rate": 4.68831615120275e-05,
+      "loss": 2.4182,
+      "step": 11801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8129996061325073,
+      "learning_rate": 4.6882897171556965e-05,
+      "loss": 1.467,
+      "step": 11802
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7215064764022827,
+      "learning_rate": 4.6882632831086446e-05,
+      "loss": 1.4977,
+      "step": 11803
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6478638648986816,
+      "learning_rate": 4.688236849061591e-05,
+      "loss": 1.8487,
+      "step": 11804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2403345108032227,
+      "learning_rate": 4.688210415014539e-05,
+      "loss": 1.5652,
+      "step": 11805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9570709466934204,
+      "learning_rate": 4.6881839809674866e-05,
+      "loss": 1.4422,
+      "step": 11806
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.625462532043457,
+      "learning_rate": 4.688157546920433e-05,
+      "loss": 1.4202,
+      "step": 11807
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.600723385810852,
+      "learning_rate": 4.6881311128733814e-05,
+      "loss": 1.3325,
+      "step": 11808
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.752936601638794,
+      "learning_rate": 4.688104678826328e-05,
+      "loss": 1.375,
+      "step": 11809
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8216826915740967,
+      "learning_rate": 4.688078244779276e-05,
+      "loss": 1.3951,
+      "step": 11810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2526769638061523,
+      "learning_rate": 4.6880518107322234e-05,
+      "loss": 1.6108,
+      "step": 11811
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.516897439956665,
+      "learning_rate": 4.688025376685171e-05,
+      "loss": 1.2669,
+      "step": 11812
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4933815002441406,
+      "learning_rate": 4.687998942638118e-05,
+      "loss": 1.3382,
+      "step": 11813
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.860020637512207,
+      "learning_rate": 4.6879725085910655e-05,
+      "loss": 1.3244,
+      "step": 11814
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125840663909912,
+      "learning_rate": 4.687946074544013e-05,
+      "loss": 1.5791,
+      "step": 11815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3115241527557373,
+      "learning_rate": 4.68791964049696e-05,
+      "loss": 1.514,
+      "step": 11816
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.758663296699524,
+      "learning_rate": 4.6878932064499076e-05,
+      "loss": 1.4436,
+      "step": 11817
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.453428864479065,
+      "learning_rate": 4.687866772402855e-05,
+      "loss": 1.8631,
+      "step": 11818
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6298259496688843,
+      "learning_rate": 4.687840338355802e-05,
+      "loss": 1.4063,
+      "step": 11819
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9720664024353027,
+      "learning_rate": 4.6878139043087496e-05,
+      "loss": 1.4595,
+      "step": 11820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0630719661712646,
+      "learning_rate": 4.687787470261698e-05,
+      "loss": 1.269,
+      "step": 11821
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9982768297195435,
+      "learning_rate": 4.6877610362146444e-05,
+      "loss": 1.3723,
+      "step": 11822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5729737281799316,
+      "learning_rate": 4.6877346021675924e-05,
+      "loss": 1.3967,
+      "step": 11823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4153871536254883,
+      "learning_rate": 4.687708168120539e-05,
+      "loss": 1.2789,
+      "step": 11824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6372374296188354,
+      "learning_rate": 4.687681734073487e-05,
+      "loss": 1.339,
+      "step": 11825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8989938497543335,
+      "learning_rate": 4.6876553000264345e-05,
+      "loss": 1.4717,
+      "step": 11826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2687981128692627,
+      "learning_rate": 4.687628865979382e-05,
+      "loss": 1.4418,
+      "step": 11827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8359256982803345,
+      "learning_rate": 4.687602431932329e-05,
+      "loss": 1.3769,
+      "step": 11828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4562547206878662,
+      "learning_rate": 4.6875759978852765e-05,
+      "loss": 1.3631,
+      "step": 11829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7903974056243896,
+      "learning_rate": 4.687549563838224e-05,
+      "loss": 1.3732,
+      "step": 11830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.048314094543457,
+      "learning_rate": 4.687523129791171e-05,
+      "loss": 1.5077,
+      "step": 11831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9540414810180664,
+      "learning_rate": 4.6874966957441186e-05,
+      "loss": 1.5044,
+      "step": 11832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7606877088546753,
+      "learning_rate": 4.687470261697066e-05,
+      "loss": 1.4204,
+      "step": 11833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.071819543838501,
+      "learning_rate": 4.687443827650013e-05,
+      "loss": 1.4478,
+      "step": 11834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6126716136932373,
+      "learning_rate": 4.687417393602961e-05,
+      "loss": 1.2739,
+      "step": 11835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5635218620300293,
+      "learning_rate": 4.687390959555909e-05,
+      "loss": 1.3064,
+      "step": 11836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6414310932159424,
+      "learning_rate": 4.6873645255088554e-05,
+      "loss": 1.449,
+      "step": 11837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7456748485565186,
+      "learning_rate": 4.6873380914618034e-05,
+      "loss": 1.5129,
+      "step": 11838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7253984212875366,
+      "learning_rate": 4.68731165741475e-05,
+      "loss": 1.2341,
+      "step": 11839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3231494426727295,
+      "learning_rate": 4.687285223367698e-05,
+      "loss": 1.5626,
+      "step": 11840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.597032070159912,
+      "learning_rate": 4.6872587893206455e-05,
+      "loss": 1.4215,
+      "step": 11841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7349687814712524,
+      "learning_rate": 4.687232355273592e-05,
+      "loss": 1.4649,
+      "step": 11842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6499499082565308,
+      "learning_rate": 4.68720592122654e-05,
+      "loss": 1.4308,
+      "step": 11843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.308032274246216,
+      "learning_rate": 4.687179487179487e-05,
+      "loss": 1.5211,
+      "step": 11844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.362236261367798,
+      "learning_rate": 4.687153053132435e-05,
+      "loss": 2.296,
+      "step": 11845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.440184235572815,
+      "learning_rate": 4.687126619085382e-05,
+      "loss": 1.387,
+      "step": 11846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7475112676620483,
+      "learning_rate": 4.6871001850383296e-05,
+      "loss": 1.3093,
+      "step": 11847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.013530969619751,
+      "learning_rate": 4.687073750991277e-05,
+      "loss": 1.5292,
+      "step": 11848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7251625061035156,
+      "learning_rate": 4.6870473169442244e-05,
+      "loss": 1.3598,
+      "step": 11849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6749461889266968,
+      "learning_rate": 4.687020882897172e-05,
+      "loss": 1.361,
+      "step": 11850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7001053094863892,
+      "learning_rate": 4.686994448850119e-05,
+      "loss": 1.3713,
+      "step": 11851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6504783630371094,
+      "learning_rate": 4.6869680148030664e-05,
+      "loss": 1.4739,
+      "step": 11852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6529531478881836,
+      "learning_rate": 4.686941580756014e-05,
+      "loss": 1.326,
+      "step": 11853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9728296995162964,
+      "learning_rate": 4.686915146708961e-05,
+      "loss": 1.4946,
+      "step": 11854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3726948499679565,
+      "learning_rate": 4.6868887126619085e-05,
+      "loss": 1.3463,
+      "step": 11855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8251312971115112,
+      "learning_rate": 4.6868622786148565e-05,
+      "loss": 1.6131,
+      "step": 11856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8040491342544556,
+      "learning_rate": 4.686835844567803e-05,
+      "loss": 1.3962,
+      "step": 11857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9310632944107056,
+      "learning_rate": 4.686809410520751e-05,
+      "loss": 1.4532,
+      "step": 11858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7961077690124512,
+      "learning_rate": 4.686782976473698e-05,
+      "loss": 1.4837,
+      "step": 11859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.868869423866272,
+      "learning_rate": 4.686756542426646e-05,
+      "loss": 1.318,
+      "step": 11860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7780404090881348,
+      "learning_rate": 4.686730108379593e-05,
+      "loss": 1.4321,
+      "step": 11861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8339322805404663,
+      "learning_rate": 4.686703674332541e-05,
+      "loss": 1.4133,
+      "step": 11862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8534077405929565,
+      "learning_rate": 4.686677240285488e-05,
+      "loss": 1.5317,
+      "step": 11863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2683725357055664,
+      "learning_rate": 4.6866508062384354e-05,
+      "loss": 1.4112,
+      "step": 11864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2959203720092773,
+      "learning_rate": 4.686624372191383e-05,
+      "loss": 1.4964,
+      "step": 11865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5357662439346313,
+      "learning_rate": 4.68659793814433e-05,
+      "loss": 1.2167,
+      "step": 11866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.977048397064209,
+      "learning_rate": 4.6865715040972775e-05,
+      "loss": 1.4979,
+      "step": 11867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5346304178237915,
+      "learning_rate": 4.686545070050225e-05,
+      "loss": 1.3641,
+      "step": 11868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4579250812530518,
+      "learning_rate": 4.686518636003172e-05,
+      "loss": 1.5354,
+      "step": 11869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.013589859008789,
+      "learning_rate": 4.6864922019561195e-05,
+      "loss": 1.5029,
+      "step": 11870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9176182746887207,
+      "learning_rate": 4.6864657679090676e-05,
+      "loss": 1.4782,
+      "step": 11871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6202943325042725,
+      "learning_rate": 4.686439333862014e-05,
+      "loss": 1.4662,
+      "step": 11872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.675308108329773,
+      "learning_rate": 4.686412899814962e-05,
+      "loss": 1.4786,
+      "step": 11873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9607267379760742,
+      "learning_rate": 4.686386465767909e-05,
+      "loss": 1.4309,
+      "step": 11874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9289294481277466,
+      "learning_rate": 4.686360031720857e-05,
+      "loss": 1.4652,
+      "step": 11875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4688317775726318,
+      "learning_rate": 4.6863335976738043e-05,
+      "loss": 2.3389,
+      "step": 11876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6024253368377686,
+      "learning_rate": 4.686307163626752e-05,
+      "loss": 1.3579,
+      "step": 11877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.126368999481201,
+      "learning_rate": 4.686280729579699e-05,
+      "loss": 1.492,
+      "step": 11878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9217482805252075,
+      "learning_rate": 4.686254295532646e-05,
+      "loss": 1.4041,
+      "step": 11879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6146563291549683,
+      "learning_rate": 4.686227861485594e-05,
+      "loss": 1.4959,
+      "step": 11880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1975820064544678,
+      "learning_rate": 4.686201427438541e-05,
+      "loss": 1.2959,
+      "step": 11881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.501268982887268,
+      "learning_rate": 4.6861749933914885e-05,
+      "loss": 1.437,
+      "step": 11882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.760039210319519,
+      "learning_rate": 4.686148559344436e-05,
+      "loss": 1.4641,
+      "step": 11883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6197290420532227,
+      "learning_rate": 4.686122125297383e-05,
+      "loss": 1.3779,
+      "step": 11884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5708556175231934,
+      "learning_rate": 4.6860956912503306e-05,
+      "loss": 1.3848,
+      "step": 11885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.29365611076355,
+      "learning_rate": 4.6860692572032786e-05,
+      "loss": 1.3906,
+      "step": 11886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.143786668777466,
+      "learning_rate": 4.686042823156225e-05,
+      "loss": 1.4239,
+      "step": 11887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8758642673492432,
+      "learning_rate": 4.6860163891091726e-05,
+      "loss": 2.6038,
+      "step": 11888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3812412023544312,
+      "learning_rate": 4.68598995506212e-05,
+      "loss": 1.3157,
+      "step": 11889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6266759634017944,
+      "learning_rate": 4.6859635210150673e-05,
+      "loss": 1.3464,
+      "step": 11890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.488694190979004,
+      "learning_rate": 4.6859370869680154e-05,
+      "loss": 1.3162,
+      "step": 11891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9798861742019653,
+      "learning_rate": 4.685910652920962e-05,
+      "loss": 1.4174,
+      "step": 11892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6528735160827637,
+      "learning_rate": 4.68588421887391e-05,
+      "loss": 1.3123,
+      "step": 11893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.484911561012268,
+      "learning_rate": 4.685857784826857e-05,
+      "loss": 1.3806,
+      "step": 11894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8840479850769043,
+      "learning_rate": 4.685831350779805e-05,
+      "loss": 1.3776,
+      "step": 11895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6982730627059937,
+      "learning_rate": 4.685804916732752e-05,
+      "loss": 1.1967,
+      "step": 11896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.028384208679199,
+      "learning_rate": 4.6857784826856995e-05,
+      "loss": 1.338,
+      "step": 11897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9885984659194946,
+      "learning_rate": 4.685752048638647e-05,
+      "loss": 1.4549,
+      "step": 11898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8418798446655273,
+      "learning_rate": 4.685725614591594e-05,
+      "loss": 1.5295,
+      "step": 11899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1926233768463135,
+      "learning_rate": 4.6856991805445416e-05,
+      "loss": 1.5306,
+      "step": 11900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.061877965927124,
+      "learning_rate": 4.685672746497489e-05,
+      "loss": 1.5193,
+      "step": 11901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.310013771057129,
+      "learning_rate": 4.685646312450436e-05,
+      "loss": 1.4583,
+      "step": 11902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7631577253341675,
+      "learning_rate": 4.685619878403384e-05,
+      "loss": 1.6436,
+      "step": 11903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5801782608032227,
+      "learning_rate": 4.685593444356331e-05,
+      "loss": 1.4761,
+      "step": 11904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.201928734779358,
+      "learning_rate": 4.6855670103092784e-05,
+      "loss": 1.9271,
+      "step": 11905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8120940923690796,
+      "learning_rate": 4.6855405762622264e-05,
+      "loss": 1.4539,
+      "step": 11906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9013662338256836,
+      "learning_rate": 4.685514142215173e-05,
+      "loss": 1.4929,
+      "step": 11907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5461734533309937,
+      "learning_rate": 4.685487708168121e-05,
+      "loss": 1.3965,
+      "step": 11908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9465947151184082,
+      "learning_rate": 4.685461274121068e-05,
+      "loss": 1.2297,
+      "step": 11909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2393676042556763,
+      "learning_rate": 4.685434840074016e-05,
+      "loss": 1.3086,
+      "step": 11910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4724222421646118,
+      "learning_rate": 4.685408406026963e-05,
+      "loss": 1.2714,
+      "step": 11911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.880279779434204,
+      "learning_rate": 4.6853819719799106e-05,
+      "loss": 1.4104,
+      "step": 11912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0290379524230957,
+      "learning_rate": 4.685355537932858e-05,
+      "loss": 1.5059,
+      "step": 11913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7665127515792847,
+      "learning_rate": 4.6853291038858046e-05,
+      "loss": 1.2909,
+      "step": 11914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4181835651397705,
+      "learning_rate": 4.6853026698387526e-05,
+      "loss": 1.45,
+      "step": 11915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5796937942504883,
+      "learning_rate": 4.6852762357917e-05,
+      "loss": 1.4487,
+      "step": 11916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7874889373779297,
+      "learning_rate": 4.6852498017446473e-05,
+      "loss": 1.3848,
+      "step": 11917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7559049129486084,
+      "learning_rate": 4.685223367697595e-05,
+      "loss": 1.3665,
+      "step": 11918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8315037488937378,
+      "learning_rate": 4.685196933650542e-05,
+      "loss": 1.4836,
+      "step": 11919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6955004930496216,
+      "learning_rate": 4.6851704996034894e-05,
+      "loss": 1.3525,
+      "step": 11920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.816684365272522,
+      "learning_rate": 4.6851440655564374e-05,
+      "loss": 1.4291,
+      "step": 11921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3805195093154907,
+      "learning_rate": 4.685117631509384e-05,
+      "loss": 1.2223,
+      "step": 11922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3284125328063965,
+      "learning_rate": 4.685091197462332e-05,
+      "loss": 1.2506,
+      "step": 11923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.732303261756897,
+      "learning_rate": 4.685064763415279e-05,
+      "loss": 1.254,
+      "step": 11924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9728180170059204,
+      "learning_rate": 4.685038329368226e-05,
+      "loss": 1.177,
+      "step": 11925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.832946538925171,
+      "learning_rate": 4.685011895321174e-05,
+      "loss": 1.4284,
+      "step": 11926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.66135573387146,
+      "learning_rate": 4.684985461274121e-05,
+      "loss": 1.3857,
+      "step": 11927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.604526996612549,
+      "learning_rate": 4.684959027227069e-05,
+      "loss": 1.3419,
+      "step": 11928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7031723260879517,
+      "learning_rate": 4.6849325931800156e-05,
+      "loss": 1.4201,
+      "step": 11929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6456395387649536,
+      "learning_rate": 4.684906159132964e-05,
+      "loss": 1.4193,
+      "step": 11930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.875165581703186,
+      "learning_rate": 4.684879725085911e-05,
+      "loss": 1.5714,
+      "step": 11931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5674526691436768,
+      "learning_rate": 4.6848532910388584e-05,
+      "loss": 1.688,
+      "step": 11932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.051806688308716,
+      "learning_rate": 4.684826856991806e-05,
+      "loss": 1.3858,
+      "step": 11933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7333849668502808,
+      "learning_rate": 4.684800422944753e-05,
+      "loss": 1.3295,
+      "step": 11934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.118396282196045,
+      "learning_rate": 4.6847739888977004e-05,
+      "loss": 1.421,
+      "step": 11935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5667097568511963,
+      "learning_rate": 4.684747554850648e-05,
+      "loss": 1.5766,
+      "step": 11936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9514882564544678,
+      "learning_rate": 4.684721120803595e-05,
+      "loss": 1.7527,
+      "step": 11937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3456331491470337,
+      "learning_rate": 4.6846946867565425e-05,
+      "loss": 1.2151,
+      "step": 11938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6891881227493286,
+      "learning_rate": 4.68466825270949e-05,
+      "loss": 1.3528,
+      "step": 11939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6078370809555054,
+      "learning_rate": 4.684641818662437e-05,
+      "loss": 1.272,
+      "step": 11940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9033845663070679,
+      "learning_rate": 4.684615384615385e-05,
+      "loss": 1.5034,
+      "step": 11941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7715864181518555,
+      "learning_rate": 4.684588950568332e-05,
+      "loss": 1.4564,
+      "step": 11942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6594271659851074,
+      "learning_rate": 4.68456251652128e-05,
+      "loss": 1.2787,
+      "step": 11943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.286046981811523,
+      "learning_rate": 4.684536082474227e-05,
+      "loss": 1.5342,
+      "step": 11944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7000694274902344,
+      "learning_rate": 4.684509648427175e-05,
+      "loss": 1.4207,
+      "step": 11945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7042996883392334,
+      "learning_rate": 4.684483214380122e-05,
+      "loss": 1.4735,
+      "step": 11946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.724159836769104,
+      "learning_rate": 4.6844567803330694e-05,
+      "loss": 1.2445,
+      "step": 11947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6746470928192139,
+      "learning_rate": 4.684430346286017e-05,
+      "loss": 1.3902,
+      "step": 11948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5380854606628418,
+      "learning_rate": 4.684403912238964e-05,
+      "loss": 1.2764,
+      "step": 11949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7545292377471924,
+      "learning_rate": 4.6843774781919115e-05,
+      "loss": 1.4562,
+      "step": 11950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8110969066619873,
+      "learning_rate": 4.684351044144859e-05,
+      "loss": 1.4079,
+      "step": 11951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.540327787399292,
+      "learning_rate": 4.684324610097806e-05,
+      "loss": 1.352,
+      "step": 11952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8552825450897217,
+      "learning_rate": 4.6842981760507536e-05,
+      "loss": 1.5655,
+      "step": 11953
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7234108448028564,
+      "learning_rate": 4.684271742003701e-05,
+      "loss": 1.4119,
+      "step": 11954
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4239702224731445,
+      "learning_rate": 4.684245307956648e-05,
+      "loss": 1.2251,
+      "step": 11955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5887788534164429,
+      "learning_rate": 4.684218873909596e-05,
+      "loss": 1.6568,
+      "step": 11956
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7679579257965088,
+      "learning_rate": 4.684192439862543e-05,
+      "loss": 1.3944,
+      "step": 11957
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3610670566558838,
+      "learning_rate": 4.684166005815491e-05,
+      "loss": 1.3583,
+      "step": 11958
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2384328842163086,
+      "learning_rate": 4.684139571768438e-05,
+      "loss": 1.4167,
+      "step": 11959
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.061800718307495,
+      "learning_rate": 4.684113137721385e-05,
+      "loss": 1.8203,
+      "step": 11960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9452909231185913,
+      "learning_rate": 4.684086703674333e-05,
+      "loss": 1.5022,
+      "step": 11961
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4950578212738037,
+      "learning_rate": 4.68406026962728e-05,
+      "loss": 1.5207,
+      "step": 11962
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0816822052001953,
+      "learning_rate": 4.684033835580228e-05,
+      "loss": 1.5987,
+      "step": 11963
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.544204592704773,
+      "learning_rate": 4.6840074015331745e-05,
+      "loss": 1.3234,
+      "step": 11964
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2983651161193848,
+      "learning_rate": 4.6839809674861225e-05,
+      "loss": 1.3092,
+      "step": 11965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.040022611618042,
+      "learning_rate": 4.68395453343907e-05,
+      "loss": 1.4101,
+      "step": 11966
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3470213413238525,
+      "learning_rate": 4.683928099392017e-05,
+      "loss": 1.4947,
+      "step": 11967
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.271844744682312,
+      "learning_rate": 4.6839016653449646e-05,
+      "loss": 2.0916,
+      "step": 11968
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.417195200920105,
+      "learning_rate": 4.683875231297912e-05,
+      "loss": 1.2418,
+      "step": 11969
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0430715084075928,
+      "learning_rate": 4.683848797250859e-05,
+      "loss": 1.4202,
+      "step": 11970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8015587329864502,
+      "learning_rate": 4.6838223632038067e-05,
+      "loss": 2.1638,
+      "step": 11971
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7381120920181274,
+      "learning_rate": 4.683795929156754e-05,
+      "loss": 1.4451,
+      "step": 11972
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6634575128555298,
+      "learning_rate": 4.6837694951097014e-05,
+      "loss": 1.3838,
+      "step": 11973
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.088592529296875,
+      "learning_rate": 4.683743061062649e-05,
+      "loss": 1.4414,
+      "step": 11974
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3598272800445557,
+      "learning_rate": 4.683716627015596e-05,
+      "loss": 1.4949,
+      "step": 11975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5087790489196777,
+      "learning_rate": 4.683690192968544e-05,
+      "loss": 1.3926,
+      "step": 11976
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.001349925994873,
+      "learning_rate": 4.683663758921491e-05,
+      "loss": 1.4704,
+      "step": 11977
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.759783148765564,
+      "learning_rate": 4.683637324874439e-05,
+      "loss": 1.4242,
+      "step": 11978
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0476129055023193,
+      "learning_rate": 4.6836108908273855e-05,
+      "loss": 1.4358,
+      "step": 11979
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2300217151641846,
+      "learning_rate": 4.6835844567803336e-05,
+      "loss": 1.4692,
+      "step": 11980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8167222738265991,
+      "learning_rate": 4.683558022733281e-05,
+      "loss": 1.2355,
+      "step": 11981
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.860203742980957,
+      "learning_rate": 4.683531588686228e-05,
+      "loss": 1.5238,
+      "step": 11982
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8649848699569702,
+      "learning_rate": 4.6835051546391756e-05,
+      "loss": 1.2959,
+      "step": 11983
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1751255989074707,
+      "learning_rate": 4.683478720592123e-05,
+      "loss": 1.4837,
+      "step": 11984
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8057743310928345,
+      "learning_rate": 4.68345228654507e-05,
+      "loss": 1.3484,
+      "step": 11985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6453022956848145,
+      "learning_rate": 4.683425852498018e-05,
+      "loss": 1.3616,
+      "step": 11986
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6764737367630005,
+      "learning_rate": 4.683399418450965e-05,
+      "loss": 1.3594,
+      "step": 11987
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7009384632110596,
+      "learning_rate": 4.6833729844039124e-05,
+      "loss": 1.4852,
+      "step": 11988
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8848950862884521,
+      "learning_rate": 4.68334655035686e-05,
+      "loss": 1.509,
+      "step": 11989
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0666003227233887,
+      "learning_rate": 4.683320116309807e-05,
+      "loss": 1.5461,
+      "step": 11990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.800054907798767,
+      "learning_rate": 4.683293682262755e-05,
+      "loss": 1.4859,
+      "step": 11991
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9320869445800781,
+      "learning_rate": 4.683267248215702e-05,
+      "loss": 1.3956,
+      "step": 11992
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.910514235496521,
+      "learning_rate": 4.68324081416865e-05,
+      "loss": 1.3885,
+      "step": 11993
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.006068706512451,
+      "learning_rate": 4.6832143801215965e-05,
+      "loss": 1.5267,
+      "step": 11994
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6025948524475098,
+      "learning_rate": 4.6831879460745446e-05,
+      "loss": 1.323,
+      "step": 11995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.695491909980774,
+      "learning_rate": 4.683161512027492e-05,
+      "loss": 1.5208,
+      "step": 11996
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.914138913154602,
+      "learning_rate": 4.6831350779804386e-05,
+      "loss": 1.5104,
+      "step": 11997
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6119409799575806,
+      "learning_rate": 4.6831086439333867e-05,
+      "loss": 1.299,
+      "step": 11998
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.230971336364746,
+      "learning_rate": 4.683082209886333e-05,
+      "loss": 1.4224,
+      "step": 11999
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.469441294670105,
+      "learning_rate": 4.6830557758392814e-05,
+      "loss": 1.3593,
+      "step": 12000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3276053667068481,
+      "learning_rate": 4.683029341792229e-05,
+      "loss": 1.5232,
+      "step": 12001
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6432833671569824,
+      "learning_rate": 4.683002907745176e-05,
+      "loss": 2.5883,
+      "step": 12002
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6686652898788452,
+      "learning_rate": 4.6829764736981234e-05,
+      "loss": 1.2806,
+      "step": 12003
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5953985452651978,
+      "learning_rate": 4.682950039651071e-05,
+      "loss": 1.3957,
+      "step": 12004
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.908709168434143,
+      "learning_rate": 4.682923605604018e-05,
+      "loss": 1.4139,
+      "step": 12005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3463246822357178,
+      "learning_rate": 4.6828971715569655e-05,
+      "loss": 1.3793,
+      "step": 12006
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.160521984100342,
+      "learning_rate": 4.682870737509913e-05,
+      "loss": 1.5302,
+      "step": 12007
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9470354318618774,
+      "learning_rate": 4.68284430346286e-05,
+      "loss": 1.521,
+      "step": 12008
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7175469398498535,
+      "learning_rate": 4.6828178694158076e-05,
+      "loss": 1.3719,
+      "step": 12009
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.22192645072937,
+      "learning_rate": 4.682791435368755e-05,
+      "loss": 1.4225,
+      "step": 12010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6197110414505005,
+      "learning_rate": 4.682765001321703e-05,
+      "loss": 1.3657,
+      "step": 12011
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5861753225326538,
+      "learning_rate": 4.6827385672746497e-05,
+      "loss": 1.4309,
+      "step": 12012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8092868328094482,
+      "learning_rate": 4.682712133227598e-05,
+      "loss": 1.3391,
+      "step": 12013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7206588983535767,
+      "learning_rate": 4.6826856991805444e-05,
+      "loss": 1.4086,
+      "step": 12014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6393331289291382,
+      "learning_rate": 4.6826592651334924e-05,
+      "loss": 1.3676,
+      "step": 12015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6485412120819092,
+      "learning_rate": 4.68263283108644e-05,
+      "loss": 1.3452,
+      "step": 12016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.63205885887146,
+      "learning_rate": 4.682606397039387e-05,
+      "loss": 1.4989,
+      "step": 12017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6545710563659668,
+      "learning_rate": 4.6825799629923345e-05,
+      "loss": 1.3646,
+      "step": 12018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.752370834350586,
+      "learning_rate": 4.682553528945282e-05,
+      "loss": 1.3536,
+      "step": 12019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2730547189712524,
+      "learning_rate": 4.682527094898229e-05,
+      "loss": 1.3542,
+      "step": 12020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3930168151855469,
+      "learning_rate": 4.6825006608511765e-05,
+      "loss": 1.4185,
+      "step": 12021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.932277798652649,
+      "learning_rate": 4.682474226804124e-05,
+      "loss": 1.387,
+      "step": 12022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.546837568283081,
+      "learning_rate": 4.682447792757071e-05,
+      "loss": 1.2784,
+      "step": 12023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2870248556137085,
+      "learning_rate": 4.6824213587100186e-05,
+      "loss": 1.2458,
+      "step": 12024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7165594100952148,
+      "learning_rate": 4.682394924662966e-05,
+      "loss": 1.4524,
+      "step": 12025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.412074089050293,
+      "learning_rate": 4.682368490615914e-05,
+      "loss": 1.7313,
+      "step": 12026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.349259376525879,
+      "learning_rate": 4.682342056568861e-05,
+      "loss": 1.3566,
+      "step": 12027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.049234628677368,
+      "learning_rate": 4.682315622521809e-05,
+      "loss": 1.516,
+      "step": 12028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9335696697235107,
+      "learning_rate": 4.6822891884747554e-05,
+      "loss": 1.5672,
+      "step": 12029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.631710410118103,
+      "learning_rate": 4.6822627544277034e-05,
+      "loss": 1.4102,
+      "step": 12030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8345474004745483,
+      "learning_rate": 4.682236320380651e-05,
+      "loss": 1.3116,
+      "step": 12031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9584441184997559,
+      "learning_rate": 4.6822098863335975e-05,
+      "loss": 2.5969,
+      "step": 12032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.681467890739441,
+      "learning_rate": 4.6821834522865455e-05,
+      "loss": 1.4713,
+      "step": 12033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8534730672836304,
+      "learning_rate": 4.682157018239492e-05,
+      "loss": 1.3843,
+      "step": 12034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8053330183029175,
+      "learning_rate": 4.68213058419244e-05,
+      "loss": 1.3907,
+      "step": 12035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6388254165649414,
+      "learning_rate": 4.6821041501453876e-05,
+      "loss": 1.405,
+      "step": 12036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.825526237487793,
+      "learning_rate": 4.682077716098335e-05,
+      "loss": 1.1637,
+      "step": 12037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.669389486312866,
+      "learning_rate": 4.682051282051282e-05,
+      "loss": 1.4919,
+      "step": 12038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.930812120437622,
+      "learning_rate": 4.6820248480042297e-05,
+      "loss": 1.5227,
+      "step": 12039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7077200412750244,
+      "learning_rate": 4.681998413957177e-05,
+      "loss": 1.2264,
+      "step": 12040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.393729567527771,
+      "learning_rate": 4.681971979910125e-05,
+      "loss": 1.2485,
+      "step": 12041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8881436586380005,
+      "learning_rate": 4.681945545863072e-05,
+      "loss": 1.488,
+      "step": 12042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.606961965560913,
+      "learning_rate": 4.681919111816019e-05,
+      "loss": 1.2548,
+      "step": 12043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.430585861206055,
+      "learning_rate": 4.6818926777689664e-05,
+      "loss": 1.3844,
+      "step": 12044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.676822304725647,
+      "learning_rate": 4.681866243721914e-05,
+      "loss": 1.3771,
+      "step": 12045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7771451473236084,
+      "learning_rate": 4.681839809674862e-05,
+      "loss": 1.2525,
+      "step": 12046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1064274311065674,
+      "learning_rate": 4.6818133756278085e-05,
+      "loss": 1.3602,
+      "step": 12047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6197340488433838,
+      "learning_rate": 4.6817869415807565e-05,
+      "loss": 1.2451,
+      "step": 12048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7432663440704346,
+      "learning_rate": 4.681760507533703e-05,
+      "loss": 1.4365,
+      "step": 12049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9380507469177246,
+      "learning_rate": 4.681734073486651e-05,
+      "loss": 1.3651,
+      "step": 12050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.432223081588745,
+      "learning_rate": 4.6817076394395986e-05,
+      "loss": 1.6031,
+      "step": 12051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9191675186157227,
+      "learning_rate": 4.681681205392546e-05,
+      "loss": 1.3977,
+      "step": 12052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.152596950531006,
+      "learning_rate": 4.681654771345493e-05,
+      "loss": 1.4314,
+      "step": 12053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5220750570297241,
+      "learning_rate": 4.681628337298441e-05,
+      "loss": 1.2849,
+      "step": 12054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.908333420753479,
+      "learning_rate": 4.681601903251388e-05,
+      "loss": 1.5112,
+      "step": 12055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7657418251037598,
+      "learning_rate": 4.6815754692043354e-05,
+      "loss": 1.4576,
+      "step": 12056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4514778852462769,
+      "learning_rate": 4.681549035157283e-05,
+      "loss": 1.4127,
+      "step": 12057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7274788618087769,
+      "learning_rate": 4.68152260111023e-05,
+      "loss": 1.4,
+      "step": 12058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.511466383934021,
+      "learning_rate": 4.6814961670631775e-05,
+      "loss": 1.3332,
+      "step": 12059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.689434289932251,
+      "learning_rate": 4.681469733016125e-05,
+      "loss": 1.4349,
+      "step": 12060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.060812473297119,
+      "learning_rate": 4.681443298969073e-05,
+      "loss": 1.5098,
+      "step": 12061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.90295147895813,
+      "learning_rate": 4.6814168649220195e-05,
+      "loss": 1.5676,
+      "step": 12062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9006764888763428,
+      "learning_rate": 4.6813904308749676e-05,
+      "loss": 1.3794,
+      "step": 12063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7517462968826294,
+      "learning_rate": 4.681363996827914e-05,
+      "loss": 1.8415,
+      "step": 12064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4389123916625977,
+      "learning_rate": 4.681337562780862e-05,
+      "loss": 1.1629,
+      "step": 12065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7660930156707764,
+      "learning_rate": 4.6813111287338096e-05,
+      "loss": 1.2398,
+      "step": 12066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4243113994598389,
+      "learning_rate": 4.681284694686757e-05,
+      "loss": 1.3177,
+      "step": 12067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4546723365783691,
+      "learning_rate": 4.6812582606397044e-05,
+      "loss": 1.2553,
+      "step": 12068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1293210983276367,
+      "learning_rate": 4.681231826592651e-05,
+      "loss": 1.4043,
+      "step": 12069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8732315301895142,
+      "learning_rate": 4.681205392545599e-05,
+      "loss": 1.3923,
+      "step": 12070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9350968599319458,
+      "learning_rate": 4.6811789584985464e-05,
+      "loss": 1.6032,
+      "step": 12071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0934810638427734,
+      "learning_rate": 4.681152524451494e-05,
+      "loss": 1.3403,
+      "step": 12072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9894907474517822,
+      "learning_rate": 4.681126090404441e-05,
+      "loss": 1.4686,
+      "step": 12073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9656070470809937,
+      "learning_rate": 4.6810996563573885e-05,
+      "loss": 1.3981,
+      "step": 12074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1371002197265625,
+      "learning_rate": 4.681073222310336e-05,
+      "loss": 1.5414,
+      "step": 12075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1993560791015625,
+      "learning_rate": 4.681046788263284e-05,
+      "loss": 1.5562,
+      "step": 12076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8344324827194214,
+      "learning_rate": 4.6810203542162306e-05,
+      "loss": 1.4224,
+      "step": 12077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4250895977020264,
+      "learning_rate": 4.680993920169178e-05,
+      "loss": 1.1588,
+      "step": 12078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5610427856445312,
+      "learning_rate": 4.680967486122125e-05,
+      "loss": 1.3825,
+      "step": 12079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6269054412841797,
+      "learning_rate": 4.6809410520750726e-05,
+      "loss": 1.5633,
+      "step": 12080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6384156942367554,
+      "learning_rate": 4.680914618028021e-05,
+      "loss": 1.7266,
+      "step": 12081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5106496810913086,
+      "learning_rate": 4.6808881839809674e-05,
+      "loss": 1.3552,
+      "step": 12082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.913558006286621,
+      "learning_rate": 4.6808617499339154e-05,
+      "loss": 1.4983,
+      "step": 12083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.774573564529419,
+      "learning_rate": 4.680835315886862e-05,
+      "loss": 1.3412,
+      "step": 12084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6512402296066284,
+      "learning_rate": 4.68080888183981e-05,
+      "loss": 1.4315,
+      "step": 12085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.814762830734253,
+      "learning_rate": 4.6807824477927575e-05,
+      "loss": 1.3783,
+      "step": 12086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6673729419708252,
+      "learning_rate": 4.680756013745705e-05,
+      "loss": 1.4218,
+      "step": 12087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6224702596664429,
+      "learning_rate": 4.680729579698652e-05,
+      "loss": 1.4449,
+      "step": 12088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.570218563079834,
+      "learning_rate": 4.6807031456515995e-05,
+      "loss": 1.4118,
+      "step": 12089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5343034267425537,
+      "learning_rate": 4.680676711604547e-05,
+      "loss": 1.2322,
+      "step": 12090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7156825065612793,
+      "learning_rate": 4.680650277557494e-05,
+      "loss": 1.487,
+      "step": 12091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.352586269378662,
+      "learning_rate": 4.6806238435104416e-05,
+      "loss": 1.2516,
+      "step": 12092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7747758626937866,
+      "learning_rate": 4.680597409463389e-05,
+      "loss": 2.0062,
+      "step": 12093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2199127674102783,
+      "learning_rate": 4.680570975416336e-05,
+      "loss": 1.6063,
+      "step": 12094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9312899112701416,
+      "learning_rate": 4.680544541369284e-05,
+      "loss": 1.5434,
+      "step": 12095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6937676668167114,
+      "learning_rate": 4.680518107322232e-05,
+      "loss": 1.3345,
+      "step": 12096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7265392541885376,
+      "learning_rate": 4.6804916732751784e-05,
+      "loss": 1.3314,
+      "step": 12097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.740363359451294,
+      "learning_rate": 4.6804652392281264e-05,
+      "loss": 1.4766,
+      "step": 12098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.754698395729065,
+      "learning_rate": 4.680438805181073e-05,
+      "loss": 1.2802,
+      "step": 12099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5406217575073242,
+      "learning_rate": 4.680412371134021e-05,
+      "loss": 1.3427,
+      "step": 12100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8995041847229004,
+      "learning_rate": 4.6803859370869685e-05,
+      "loss": 1.4589,
+      "step": 12101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3232293128967285,
+      "learning_rate": 4.680359503039916e-05,
+      "loss": 1.56,
+      "step": 12102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2404139041900635,
+      "learning_rate": 4.680333068992863e-05,
+      "loss": 1.4285,
+      "step": 12103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5108888149261475,
+      "learning_rate": 4.68030663494581e-05,
+      "loss": 1.2145,
+      "step": 12104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4873452186584473,
+      "learning_rate": 4.680280200898758e-05,
+      "loss": 1.4385,
+      "step": 12105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3241925239562988,
+      "learning_rate": 4.680253766851705e-05,
+      "loss": 1.0121,
+      "step": 12106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.758816719055176,
+      "learning_rate": 4.6802273328046526e-05,
+      "loss": 1.5556,
+      "step": 12107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8570889234542847,
+      "learning_rate": 4.6802008987576e-05,
+      "loss": 1.2539,
+      "step": 12108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7382607460021973,
+      "learning_rate": 4.6801744647105474e-05,
+      "loss": 1.4958,
+      "step": 12109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6538640260696411,
+      "learning_rate": 4.680148030663495e-05,
+      "loss": 1.343,
+      "step": 12110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6728029251098633,
+      "learning_rate": 4.680121596616443e-05,
+      "loss": 1.3386,
+      "step": 12111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0324485301971436,
+      "learning_rate": 4.6800951625693894e-05,
+      "loss": 1.4089,
+      "step": 12112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.521549701690674,
+      "learning_rate": 4.6800687285223375e-05,
+      "loss": 1.5382,
+      "step": 12113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4087978601455688,
+      "learning_rate": 4.680042294475284e-05,
+      "loss": 1.3196,
+      "step": 12114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7893182039260864,
+      "learning_rate": 4.6800158604282315e-05,
+      "loss": 1.2978,
+      "step": 12115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8886936902999878,
+      "learning_rate": 4.6799894263811795e-05,
+      "loss": 1.4731,
+      "step": 12116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4321653842926025,
+      "learning_rate": 4.679962992334126e-05,
+      "loss": 1.8686,
+      "step": 12117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3753223419189453,
+      "learning_rate": 4.679936558287074e-05,
+      "loss": 1.4874,
+      "step": 12118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6131004095077515,
+      "learning_rate": 4.679910124240021e-05,
+      "loss": 1.2187,
+      "step": 12119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8173130750656128,
+      "learning_rate": 4.679883690192969e-05,
+      "loss": 1.3062,
+      "step": 12120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8478742837905884,
+      "learning_rate": 4.679857256145916e-05,
+      "loss": 1.3936,
+      "step": 12121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.055654287338257,
+      "learning_rate": 4.679830822098864e-05,
+      "loss": 1.5004,
+      "step": 12122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0901262760162354,
+      "learning_rate": 4.679804388051811e-05,
+      "loss": 1.5877,
+      "step": 12123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.88284170627594,
+      "learning_rate": 4.6797779540047584e-05,
+      "loss": 2.723,
+      "step": 12124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7334318161010742,
+      "learning_rate": 4.679751519957706e-05,
+      "loss": 1.5476,
+      "step": 12125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0017309188842773,
+      "learning_rate": 4.679725085910653e-05,
+      "loss": 1.3443,
+      "step": 12126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3036757707595825,
+      "learning_rate": 4.6796986518636005e-05,
+      "loss": 1.1031,
+      "step": 12127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.623099446296692,
+      "learning_rate": 4.679672217816548e-05,
+      "loss": 1.3353,
+      "step": 12128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8170883655548096,
+      "learning_rate": 4.679645783769495e-05,
+      "loss": 1.4449,
+      "step": 12129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5988396406173706,
+      "learning_rate": 4.6796193497224425e-05,
+      "loss": 1.4752,
+      "step": 12130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4490195512771606,
+      "learning_rate": 4.6795929156753906e-05,
+      "loss": 1.2422,
+      "step": 12131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9427034854888916,
+      "learning_rate": 4.679566481628337e-05,
+      "loss": 1.3932,
+      "step": 12132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4290562868118286,
+      "learning_rate": 4.679540047581285e-05,
+      "loss": 1.3705,
+      "step": 12133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8716999292373657,
+      "learning_rate": 4.679513613534232e-05,
+      "loss": 1.4611,
+      "step": 12134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6684027910232544,
+      "learning_rate": 4.67948717948718e-05,
+      "loss": 1.4485,
+      "step": 12135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8493674993515015,
+      "learning_rate": 4.6794607454401274e-05,
+      "loss": 1.6047,
+      "step": 12136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3761216402053833,
+      "learning_rate": 4.679434311393075e-05,
+      "loss": 1.1603,
+      "step": 12137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.517769694328308,
+      "learning_rate": 4.679407877346022e-05,
+      "loss": 1.3676,
+      "step": 12138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5290968418121338,
+      "learning_rate": 4.6793814432989694e-05,
+      "loss": 1.4036,
+      "step": 12139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4599353075027466,
+      "learning_rate": 4.679355009251917e-05,
+      "loss": 1.384,
+      "step": 12140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5161856412887573,
+      "learning_rate": 4.679328575204864e-05,
+      "loss": 1.4261,
+      "step": 12141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5800033807754517,
+      "learning_rate": 4.6793021411578115e-05,
+      "loss": 1.2158,
+      "step": 12142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5702581405639648,
+      "learning_rate": 4.679275707110759e-05,
+      "loss": 1.347,
+      "step": 12143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6123322248458862,
+      "learning_rate": 4.679249273063706e-05,
+      "loss": 1.2683,
+      "step": 12144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9993971586227417,
+      "learning_rate": 4.6792228390166536e-05,
+      "loss": 1.2622,
+      "step": 12145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.745434284210205,
+      "learning_rate": 4.6791964049696016e-05,
+      "loss": 1.3801,
+      "step": 12146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4585299491882324,
+      "learning_rate": 4.679169970922548e-05,
+      "loss": 1.6788,
+      "step": 12147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.634270668029785,
+      "learning_rate": 4.679143536875496e-05,
+      "loss": 1.6208,
+      "step": 12148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8851217031478882,
+      "learning_rate": 4.679117102828443e-05,
+      "loss": 1.3614,
+      "step": 12149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7857847213745117,
+      "learning_rate": 4.6790906687813904e-05,
+      "loss": 1.4597,
+      "step": 12150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9908796548843384,
+      "learning_rate": 4.6790642347343384e-05,
+      "loss": 1.5076,
+      "step": 12151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6827459335327148,
+      "learning_rate": 4.679037800687285e-05,
+      "loss": 1.537,
+      "step": 12152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.732964038848877,
+      "learning_rate": 4.679011366640233e-05,
+      "loss": 1.5596,
+      "step": 12153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.752198338508606,
+      "learning_rate": 4.67898493259318e-05,
+      "loss": 1.5031,
+      "step": 12154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.38075852394104,
+      "learning_rate": 4.678958498546128e-05,
+      "loss": 1.2867,
+      "step": 12155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5521987676620483,
+      "learning_rate": 4.678932064499075e-05,
+      "loss": 1.5766,
+      "step": 12156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.514552116394043,
+      "learning_rate": 4.6789056304520225e-05,
+      "loss": 1.3352,
+      "step": 12157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4729700088500977,
+      "learning_rate": 4.67887919640497e-05,
+      "loss": 1.4238,
+      "step": 12158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.490679383277893,
+      "learning_rate": 4.678852762357917e-05,
+      "loss": 2.5734,
+      "step": 12159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8152003288269043,
+      "learning_rate": 4.6788263283108646e-05,
+      "loss": 1.2995,
+      "step": 12160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5613795518875122,
+      "learning_rate": 4.678799894263812e-05,
+      "loss": 1.5332,
+      "step": 12161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.446499824523926,
+      "learning_rate": 4.678773460216759e-05,
+      "loss": 1.5329,
+      "step": 12162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9938427209854126,
+      "learning_rate": 4.678747026169707e-05,
+      "loss": 1.3686,
+      "step": 12163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7034209966659546,
+      "learning_rate": 4.678720592122654e-05,
+      "loss": 1.3448,
+      "step": 12164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1192641258239746,
+      "learning_rate": 4.6786941580756014e-05,
+      "loss": 1.4735,
+      "step": 12165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6267684698104858,
+      "learning_rate": 4.6786677240285494e-05,
+      "loss": 1.4342,
+      "step": 12166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9216526746749878,
+      "learning_rate": 4.678641289981496e-05,
+      "loss": 1.4192,
+      "step": 12167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4310604333877563,
+      "learning_rate": 4.678614855934444e-05,
+      "loss": 1.2472,
+      "step": 12168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6533604860305786,
+      "learning_rate": 4.678588421887391e-05,
+      "loss": 1.3507,
+      "step": 12169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.563974618911743,
+      "learning_rate": 4.678561987840339e-05,
+      "loss": 1.4022,
+      "step": 12170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.412702202796936,
+      "learning_rate": 4.678535553793286e-05,
+      "loss": 1.2805,
+      "step": 12171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0471410751342773,
+      "learning_rate": 4.6785091197462336e-05,
+      "loss": 1.4141,
+      "step": 12172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3927891254425049,
+      "learning_rate": 4.678482685699181e-05,
+      "loss": 1.332,
+      "step": 12173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.221499443054199,
+      "learning_rate": 4.678456251652128e-05,
+      "loss": 1.3474,
+      "step": 12174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4986913204193115,
+      "learning_rate": 4.6784298176050756e-05,
+      "loss": 1.2913,
+      "step": 12175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.870036244392395,
+      "learning_rate": 4.678403383558023e-05,
+      "loss": 1.4896,
+      "step": 12176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.035750389099121,
+      "learning_rate": 4.6783769495109704e-05,
+      "loss": 1.4057,
+      "step": 12177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5781807899475098,
+      "learning_rate": 4.678350515463918e-05,
+      "loss": 1.2167,
+      "step": 12178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6370803117752075,
+      "learning_rate": 4.678324081416865e-05,
+      "loss": 1.3761,
+      "step": 12179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1930205821990967,
+      "learning_rate": 4.6782976473698124e-05,
+      "loss": 1.4184,
+      "step": 12180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.370927333831787,
+      "learning_rate": 4.6782712133227605e-05,
+      "loss": 1.5319,
+      "step": 12181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8180533647537231,
+      "learning_rate": 4.678244779275707e-05,
+      "loss": 1.3826,
+      "step": 12182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7272015810012817,
+      "learning_rate": 4.678218345228655e-05,
+      "loss": 1.3552,
+      "step": 12183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3290600776672363,
+      "learning_rate": 4.678191911181602e-05,
+      "loss": 1.4773,
+      "step": 12184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5426583290100098,
+      "learning_rate": 4.67816547713455e-05,
+      "loss": 1.3245,
+      "step": 12185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.78695547580719,
+      "learning_rate": 4.678139043087497e-05,
+      "loss": 1.4205,
+      "step": 12186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.26709246635437,
+      "learning_rate": 4.678112609040444e-05,
+      "loss": 1.455,
+      "step": 12187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.052546501159668,
+      "learning_rate": 4.678086174993392e-05,
+      "loss": 1.3141,
+      "step": 12188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7487560510635376,
+      "learning_rate": 4.6780597409463386e-05,
+      "loss": 1.3913,
+      "step": 12189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5242658853530884,
+      "learning_rate": 4.678033306899287e-05,
+      "loss": 1.4005,
+      "step": 12190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6069211959838867,
+      "learning_rate": 4.678006872852234e-05,
+      "loss": 1.1105,
+      "step": 12191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5591661930084229,
+      "learning_rate": 4.6779804388051814e-05,
+      "loss": 1.6063,
+      "step": 12192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.614159107208252,
+      "learning_rate": 4.677954004758129e-05,
+      "loss": 1.5394,
+      "step": 12193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7848511934280396,
+      "learning_rate": 4.677927570711076e-05,
+      "loss": 1.4252,
+      "step": 12194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5143258571624756,
+      "learning_rate": 4.6779011366640235e-05,
+      "loss": 1.2933,
+      "step": 12195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.382547616958618,
+      "learning_rate": 4.677874702616971e-05,
+      "loss": 1.4628,
+      "step": 12196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6351890563964844,
+      "learning_rate": 4.677848268569918e-05,
+      "loss": 1.4974,
+      "step": 12197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9553179740905762,
+      "learning_rate": 4.6778218345228655e-05,
+      "loss": 1.3483,
+      "step": 12198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.818225860595703,
+      "learning_rate": 4.677795400475813e-05,
+      "loss": 1.5272,
+      "step": 12199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7009308338165283,
+      "learning_rate": 4.67776896642876e-05,
+      "loss": 1.3055,
+      "step": 12200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9276769161224365,
+      "learning_rate": 4.677742532381708e-05,
+      "loss": 1.362,
+      "step": 12201
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.970616102218628,
+      "learning_rate": 4.677716098334655e-05,
+      "loss": 1.5149,
+      "step": 12202
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7595406770706177,
+      "learning_rate": 4.677689664287603e-05,
+      "loss": 1.4113,
+      "step": 12203
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9333579540252686,
+      "learning_rate": 4.67766323024055e-05,
+      "loss": 1.4796,
+      "step": 12204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8210986852645874,
+      "learning_rate": 4.677636796193498e-05,
+      "loss": 1.3798,
+      "step": 12205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.477149248123169,
+      "learning_rate": 4.677610362146445e-05,
+      "loss": 1.4049,
+      "step": 12206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2698848247528076,
+      "learning_rate": 4.6775839280993924e-05,
+      "loss": 1.5211,
+      "step": 12207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9111337661743164,
+      "learning_rate": 4.67755749405234e-05,
+      "loss": 1.503,
+      "step": 12208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7177491188049316,
+      "learning_rate": 4.677531060005287e-05,
+      "loss": 2.444,
+      "step": 12209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.388062000274658,
+      "learning_rate": 4.6775046259582345e-05,
+      "loss": 1.5327,
+      "step": 12210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5272730588912964,
+      "learning_rate": 4.677478191911182e-05,
+      "loss": 2.2732,
+      "step": 12211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.57919979095459,
+      "learning_rate": 4.677451757864129e-05,
+      "loss": 1.4556,
+      "step": 12212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.921301245689392,
+      "learning_rate": 4.6774253238170766e-05,
+      "loss": 2.0008,
+      "step": 12213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8217763900756836,
+      "learning_rate": 4.677398889770024e-05,
+      "loss": 1.3942,
+      "step": 12214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6639338731765747,
+      "learning_rate": 4.677372455722971e-05,
+      "loss": 1.278,
+      "step": 12215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7769899368286133,
+      "learning_rate": 4.677346021675919e-05,
+      "loss": 1.3967,
+      "step": 12216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9553955793380737,
+      "learning_rate": 4.677319587628866e-05,
+      "loss": 1.3565,
+      "step": 12217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5942085981369019,
+      "learning_rate": 4.677293153581814e-05,
+      "loss": 1.5202,
+      "step": 12218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0549025535583496,
+      "learning_rate": 4.677266719534761e-05,
+      "loss": 1.2481,
+      "step": 12219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0627191066741943,
+      "learning_rate": 4.677240285487709e-05,
+      "loss": 1.5696,
+      "step": 12220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6399083137512207,
+      "learning_rate": 4.677213851440656e-05,
+      "loss": 1.2763,
+      "step": 12221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.744254469871521,
+      "learning_rate": 4.677187417393603e-05,
+      "loss": 1.3875,
+      "step": 12222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7865267992019653,
+      "learning_rate": 4.677160983346551e-05,
+      "loss": 1.4381,
+      "step": 12223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6068828105926514,
+      "learning_rate": 4.6771345492994975e-05,
+      "loss": 1.4291,
+      "step": 12224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6044600009918213,
+      "learning_rate": 4.6771081152524455e-05,
+      "loss": 1.5151,
+      "step": 12225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9979280233383179,
+      "learning_rate": 4.677081681205393e-05,
+      "loss": 1.4751,
+      "step": 12226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8663870096206665,
+      "learning_rate": 4.67705524715834e-05,
+      "loss": 1.3738,
+      "step": 12227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6429319381713867,
+      "learning_rate": 4.6770288131112876e-05,
+      "loss": 1.4697,
+      "step": 12228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.347306489944458,
+      "learning_rate": 4.677002379064235e-05,
+      "loss": 1.3047,
+      "step": 12229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7056015729904175,
+      "learning_rate": 4.676975945017182e-05,
+      "loss": 1.215,
+      "step": 12230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8123619556427002,
+      "learning_rate": 4.6769495109701303e-05,
+      "loss": 1.4869,
+      "step": 12231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6977245807647705,
+      "learning_rate": 4.676923076923077e-05,
+      "loss": 1.5897,
+      "step": 12232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.811293363571167,
+      "learning_rate": 4.6768966428760244e-05,
+      "loss": 1.3061,
+      "step": 12233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7261484861373901,
+      "learning_rate": 4.676870208828972e-05,
+      "loss": 1.5,
+      "step": 12234
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.891760230064392,
+      "learning_rate": 4.676843774781919e-05,
+      "loss": 1.3465,
+      "step": 12235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8704500198364258,
+      "learning_rate": 4.676817340734867e-05,
+      "loss": 1.3761,
+      "step": 12236
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6053502559661865,
+      "learning_rate": 4.676790906687814e-05,
+      "loss": 1.3955,
+      "step": 12237
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7845290899276733,
+      "learning_rate": 4.676764472640762e-05,
+      "loss": 1.4561,
+      "step": 12238
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.323852300643921,
+      "learning_rate": 4.6767380385937085e-05,
+      "loss": 1.3701,
+      "step": 12239
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7047414779663086,
+      "learning_rate": 4.6767116045466566e-05,
+      "loss": 1.3538,
+      "step": 12240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3010605573654175,
+      "learning_rate": 4.676685170499604e-05,
+      "loss": 1.2843,
+      "step": 12241
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8523005247116089,
+      "learning_rate": 4.676658736452551e-05,
+      "loss": 1.2914,
+      "step": 12242
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6684545278549194,
+      "learning_rate": 4.6766323024054986e-05,
+      "loss": 1.3992,
+      "step": 12243
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7254555225372314,
+      "learning_rate": 4.676605868358446e-05,
+      "loss": 1.4899,
+      "step": 12244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6880152225494385,
+      "learning_rate": 4.6765794343113933e-05,
+      "loss": 1.3527,
+      "step": 12245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0014991760253906,
+      "learning_rate": 4.676553000264341e-05,
+      "loss": 1.511,
+      "step": 12246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6400736570358276,
+      "learning_rate": 4.676526566217288e-05,
+      "loss": 1.9618,
+      "step": 12247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2852455377578735,
+      "learning_rate": 4.6765001321702354e-05,
+      "loss": 2.2189,
+      "step": 12248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9326105117797852,
+      "learning_rate": 4.676473698123183e-05,
+      "loss": 1.4447,
+      "step": 12249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.921304702758789,
+      "learning_rate": 4.67644726407613e-05,
+      "loss": 1.5973,
+      "step": 12250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9742964506149292,
+      "learning_rate": 4.676420830029078e-05,
+      "loss": 1.4313,
+      "step": 12251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.588417649269104,
+      "learning_rate": 4.676394395982025e-05,
+      "loss": 1.4382,
+      "step": 12252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6491103172302246,
+      "learning_rate": 4.676367961934973e-05,
+      "loss": 1.336,
+      "step": 12253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.087625503540039,
+      "learning_rate": 4.6763415278879196e-05,
+      "loss": 1.7077,
+      "step": 12254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.689406394958496,
+      "learning_rate": 4.6763150938408676e-05,
+      "loss": 1.3141,
+      "step": 12255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5303702354431152,
+      "learning_rate": 4.676288659793815e-05,
+      "loss": 1.3935,
+      "step": 12256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7230525016784668,
+      "learning_rate": 4.676262225746762e-05,
+      "loss": 1.3425,
+      "step": 12257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5160510540008545,
+      "learning_rate": 4.67623579169971e-05,
+      "loss": 1.3942,
+      "step": 12258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5969021320343018,
+      "learning_rate": 4.6762093576526563e-05,
+      "loss": 2.4377,
+      "step": 12259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5317317247390747,
+      "learning_rate": 4.6761829236056044e-05,
+      "loss": 1.323,
+      "step": 12260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6930686235427856,
+      "learning_rate": 4.676156489558552e-05,
+      "loss": 2.1031,
+      "step": 12261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.77438223361969,
+      "learning_rate": 4.676130055511499e-05,
+      "loss": 1.2438,
+      "step": 12262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0819404125213623,
+      "learning_rate": 4.6761036214644464e-05,
+      "loss": 1.4536,
+      "step": 12263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0331368446350098,
+      "learning_rate": 4.676077187417394e-05,
+      "loss": 1.2741,
+      "step": 12264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6578702926635742,
+      "learning_rate": 4.676050753370341e-05,
+      "loss": 1.3756,
+      "step": 12265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.762544870376587,
+      "learning_rate": 4.676024319323289e-05,
+      "loss": 1.3514,
+      "step": 12266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.856313943862915,
+      "learning_rate": 4.675997885276236e-05,
+      "loss": 1.3001,
+      "step": 12267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6293376684188843,
+      "learning_rate": 4.675971451229183e-05,
+      "loss": 1.3358,
+      "step": 12268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.044739246368408,
+      "learning_rate": 4.6759450171821306e-05,
+      "loss": 1.4645,
+      "step": 12269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6915888786315918,
+      "learning_rate": 4.675918583135078e-05,
+      "loss": 1.2978,
+      "step": 12270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9362366199493408,
+      "learning_rate": 4.675892149088026e-05,
+      "loss": 1.3226,
+      "step": 12271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6595444679260254,
+      "learning_rate": 4.675865715040973e-05,
+      "loss": 1.5319,
+      "step": 12272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7599231004714966,
+      "learning_rate": 4.675839280993921e-05,
+      "loss": 1.2192,
+      "step": 12273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5623456239700317,
+      "learning_rate": 4.6758128469468674e-05,
+      "loss": 1.4295,
+      "step": 12274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8335092067718506,
+      "learning_rate": 4.6757864128998154e-05,
+      "loss": 1.3615,
+      "step": 12275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2696967124938965,
+      "learning_rate": 4.675759978852763e-05,
+      "loss": 1.2555,
+      "step": 12276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.720158576965332,
+      "learning_rate": 4.67573354480571e-05,
+      "loss": 1.4646,
+      "step": 12277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8711953163146973,
+      "learning_rate": 4.6757071107586575e-05,
+      "loss": 1.4346,
+      "step": 12278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.738114833831787,
+      "learning_rate": 4.675680676711605e-05,
+      "loss": 1.499,
+      "step": 12279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7912750244140625,
+      "learning_rate": 4.675654242664552e-05,
+      "loss": 1.4686,
+      "step": 12280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4630610942840576,
+      "learning_rate": 4.6756278086174996e-05,
+      "loss": 1.5568,
+      "step": 12281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.648078441619873,
+      "learning_rate": 4.675601374570447e-05,
+      "loss": 1.2343,
+      "step": 12282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6418735980987549,
+      "learning_rate": 4.675574940523394e-05,
+      "loss": 1.3435,
+      "step": 12283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3261983394622803,
+      "learning_rate": 4.6755485064763416e-05,
+      "loss": 1.4462,
+      "step": 12284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.604938268661499,
+      "learning_rate": 4.675522072429289e-05,
+      "loss": 1.4097,
+      "step": 12285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8092879056930542,
+      "learning_rate": 4.675495638382237e-05,
+      "loss": 1.4839,
+      "step": 12286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6161240339279175,
+      "learning_rate": 4.675469204335184e-05,
+      "loss": 1.3702,
+      "step": 12287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0470802783966064,
+      "learning_rate": 4.675442770288132e-05,
+      "loss": 1.4149,
+      "step": 12288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8808708190917969,
+      "learning_rate": 4.6754163362410784e-05,
+      "loss": 1.4793,
+      "step": 12289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2611746788024902,
+      "learning_rate": 4.6753899021940264e-05,
+      "loss": 1.5294,
+      "step": 12290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5805155038833618,
+      "learning_rate": 4.675363468146974e-05,
+      "loss": 1.4179,
+      "step": 12291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.176103353500366,
+      "learning_rate": 4.675337034099921e-05,
+      "loss": 1.4911,
+      "step": 12292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6625422239303589,
+      "learning_rate": 4.6753106000528685e-05,
+      "loss": 1.4819,
+      "step": 12293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6637550592422485,
+      "learning_rate": 4.675284166005816e-05,
+      "loss": 1.2481,
+      "step": 12294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9638785123825073,
+      "learning_rate": 4.675257731958763e-05,
+      "loss": 1.4245,
+      "step": 12295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6666196584701538,
+      "learning_rate": 4.6752312979117106e-05,
+      "loss": 1.3359,
+      "step": 12296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8307148218154907,
+      "learning_rate": 4.675204863864658e-05,
+      "loss": 1.4029,
+      "step": 12297
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9106765985488892,
+      "learning_rate": 4.675178429817605e-05,
+      "loss": 1.3826,
+      "step": 12298
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8423395156860352,
+      "learning_rate": 4.6751519957705527e-05,
+      "loss": 1.4339,
+      "step": 12299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.314579486846924,
+      "learning_rate": 4.6751255617235e-05,
+      "loss": 1.1933,
+      "step": 12300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.929434895515442,
+      "learning_rate": 4.675099127676448e-05,
+      "loss": 1.4215,
+      "step": 12301
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6230219602584839,
+      "learning_rate": 4.675072693629395e-05,
+      "loss": 2.2314,
+      "step": 12302
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.158564567565918,
+      "learning_rate": 4.675046259582343e-05,
+      "loss": 1.4227,
+      "step": 12303
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9002760648727417,
+      "learning_rate": 4.6750198255352894e-05,
+      "loss": 1.378,
+      "step": 12304
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.482314944267273,
+      "learning_rate": 4.674993391488237e-05,
+      "loss": 1.221,
+      "step": 12305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.423277735710144,
+      "learning_rate": 4.674966957441185e-05,
+      "loss": 1.2532,
+      "step": 12306
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0372798442840576,
+      "learning_rate": 4.6749405233941315e-05,
+      "loss": 1.6141,
+      "step": 12307
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.989123821258545,
+      "learning_rate": 4.6749140893470796e-05,
+      "loss": 1.4198,
+      "step": 12308
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3023884296417236,
+      "learning_rate": 4.674887655300026e-05,
+      "loss": 1.5775,
+      "step": 12309
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9880605936050415,
+      "learning_rate": 4.674861221252974e-05,
+      "loss": 1.4478,
+      "step": 12310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5440714359283447,
+      "learning_rate": 4.6748347872059216e-05,
+      "loss": 1.2282,
+      "step": 12311
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7972569465637207,
+      "learning_rate": 4.674808353158869e-05,
+      "loss": 1.5775,
+      "step": 12312
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.30136775970459,
+      "learning_rate": 4.674781919111816e-05,
+      "loss": 1.3254,
+      "step": 12313
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9724980592727661,
+      "learning_rate": 4.674755485064764e-05,
+      "loss": 1.441,
+      "step": 12314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8876794576644897,
+      "learning_rate": 4.674729051017711e-05,
+      "loss": 1.4615,
+      "step": 12315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6743582487106323,
+      "learning_rate": 4.6747026169706584e-05,
+      "loss": 1.477,
+      "step": 12316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7142313718795776,
+      "learning_rate": 4.674676182923606e-05,
+      "loss": 1.3765,
+      "step": 12317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8281047344207764,
+      "learning_rate": 4.674649748876553e-05,
+      "loss": 1.4068,
+      "step": 12318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6907888650894165,
+      "learning_rate": 4.6746233148295005e-05,
+      "loss": 1.3106,
+      "step": 12319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.891422986984253,
+      "learning_rate": 4.674596880782448e-05,
+      "loss": 1.4181,
+      "step": 12320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0251495838165283,
+      "learning_rate": 4.674570446735396e-05,
+      "loss": 1.3473,
+      "step": 12321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7263752222061157,
+      "learning_rate": 4.6745440126883426e-05,
+      "loss": 1.3291,
+      "step": 12322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.761207103729248,
+      "learning_rate": 4.6745175786412906e-05,
+      "loss": 1.3246,
+      "step": 12323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9387636184692383,
+      "learning_rate": 4.674491144594237e-05,
+      "loss": 1.4074,
+      "step": 12324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8708958625793457,
+      "learning_rate": 4.674464710547185e-05,
+      "loss": 1.5774,
+      "step": 12325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7778946161270142,
+      "learning_rate": 4.6744382765001327e-05,
+      "loss": 1.3885,
+      "step": 12326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6515021324157715,
+      "learning_rate": 4.67441184245308e-05,
+      "loss": 1.3333,
+      "step": 12327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4682527780532837,
+      "learning_rate": 4.6743854084060274e-05,
+      "loss": 1.376,
+      "step": 12328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.07907772064209,
+      "learning_rate": 4.674358974358975e-05,
+      "loss": 1.5181,
+      "step": 12329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8137989044189453,
+      "learning_rate": 4.674332540311922e-05,
+      "loss": 1.3926,
+      "step": 12330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8650132417678833,
+      "learning_rate": 4.6743061062648694e-05,
+      "loss": 1.4283,
+      "step": 12331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2614517211914062,
+      "learning_rate": 4.674279672217817e-05,
+      "loss": 1.4374,
+      "step": 12332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.886409044265747,
+      "learning_rate": 4.674253238170764e-05,
+      "loss": 1.4073,
+      "step": 12333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9689120054244995,
+      "learning_rate": 4.6742268041237115e-05,
+      "loss": 1.4349,
+      "step": 12334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4962379932403564,
+      "learning_rate": 4.674200370076659e-05,
+      "loss": 1.3705,
+      "step": 12335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6995948553085327,
+      "learning_rate": 4.674173936029607e-05,
+      "loss": 1.3141,
+      "step": 12336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.448844313621521,
+      "learning_rate": 4.6741475019825536e-05,
+      "loss": 1.2652,
+      "step": 12337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.556330680847168,
+      "learning_rate": 4.6741210679355016e-05,
+      "loss": 1.4888,
+      "step": 12338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7648011445999146,
+      "learning_rate": 4.674094633888448e-05,
+      "loss": 1.3682,
+      "step": 12339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.639237880706787,
+      "learning_rate": 4.6740681998413957e-05,
+      "loss": 1.4556,
+      "step": 12340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.256380319595337,
+      "learning_rate": 4.674041765794344e-05,
+      "loss": 1.5168,
+      "step": 12341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4294772148132324,
+      "learning_rate": 4.6740153317472904e-05,
+      "loss": 1.2547,
+      "step": 12342
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8003883361816406,
+      "learning_rate": 4.6739888977002384e-05,
+      "loss": 1.4755,
+      "step": 12343
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8562613725662231,
+      "learning_rate": 4.673962463653185e-05,
+      "loss": 1.3275,
+      "step": 12344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4380581378936768,
+      "learning_rate": 4.673936029606133e-05,
+      "loss": 1.4186,
+      "step": 12345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7330809831619263,
+      "learning_rate": 4.6739095955590805e-05,
+      "loss": 1.3675,
+      "step": 12346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9069693088531494,
+      "learning_rate": 4.673883161512028e-05,
+      "loss": 1.5535,
+      "step": 12347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9889707565307617,
+      "learning_rate": 4.673856727464975e-05,
+      "loss": 1.3087,
+      "step": 12348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3551111221313477,
+      "learning_rate": 4.6738302934179225e-05,
+      "loss": 1.3998,
+      "step": 12349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8158022165298462,
+      "learning_rate": 4.67380385937087e-05,
+      "loss": 1.491,
+      "step": 12350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7914174795150757,
+      "learning_rate": 4.673777425323817e-05,
+      "loss": 1.4313,
+      "step": 12351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6297706365585327,
+      "learning_rate": 4.6737509912767646e-05,
+      "loss": 1.3198,
+      "step": 12352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.597848653793335,
+      "learning_rate": 4.673724557229712e-05,
+      "loss": 1.2475,
+      "step": 12353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8443225622177124,
+      "learning_rate": 4.673698123182659e-05,
+      "loss": 1.4265,
+      "step": 12354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5999486446380615,
+      "learning_rate": 4.673671689135607e-05,
+      "loss": 1.4228,
+      "step": 12355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5105069875717163,
+      "learning_rate": 4.673645255088555e-05,
+      "loss": 1.374,
+      "step": 12356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.137068510055542,
+      "learning_rate": 4.6736188210415014e-05,
+      "loss": 1.4646,
+      "step": 12357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5589375495910645,
+      "learning_rate": 4.6735923869944494e-05,
+      "loss": 1.4433,
+      "step": 12358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0023162364959717,
+      "learning_rate": 4.673565952947396e-05,
+      "loss": 1.3963,
+      "step": 12359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.679175853729248,
+      "learning_rate": 4.673539518900344e-05,
+      "loss": 2.461,
+      "step": 12360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.308805227279663,
+      "learning_rate": 4.6735130848532915e-05,
+      "loss": 1.2555,
+      "step": 12361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3240487575531006,
+      "learning_rate": 4.673486650806239e-05,
+      "loss": 1.3488,
+      "step": 12362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4236663579940796,
+      "learning_rate": 4.673460216759186e-05,
+      "loss": 1.2336,
+      "step": 12363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3419878482818604,
+      "learning_rate": 4.6734337827121336e-05,
+      "loss": 1.5723,
+      "step": 12364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7915230989456177,
+      "learning_rate": 4.673407348665081e-05,
+      "loss": 1.4572,
+      "step": 12365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.436400055885315,
+      "learning_rate": 4.673380914618028e-05,
+      "loss": 2.4181,
+      "step": 12366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9772989749908447,
+      "learning_rate": 4.6733544805709757e-05,
+      "loss": 1.476,
+      "step": 12367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.612161636352539,
+      "learning_rate": 4.673328046523923e-05,
+      "loss": 1.4256,
+      "step": 12368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.847697138786316,
+      "learning_rate": 4.6733016124768704e-05,
+      "loss": 1.3896,
+      "step": 12369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5892003774642944,
+      "learning_rate": 4.673275178429818e-05,
+      "loss": 1.4076,
+      "step": 12370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7536568641662598,
+      "learning_rate": 4.673248744382766e-05,
+      "loss": 1.3577,
+      "step": 12371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375840663909912,
+      "learning_rate": 4.6732223103357124e-05,
+      "loss": 1.6391,
+      "step": 12372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7862342596054077,
+      "learning_rate": 4.6731958762886605e-05,
+      "loss": 1.4278,
+      "step": 12373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5627858638763428,
+      "learning_rate": 4.673169442241607e-05,
+      "loss": 1.414,
+      "step": 12374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9388593435287476,
+      "learning_rate": 4.673143008194555e-05,
+      "loss": 1.3995,
+      "step": 12375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.370011806488037,
+      "learning_rate": 4.6731165741475025e-05,
+      "loss": 1.564,
+      "step": 12376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6883435249328613,
+      "learning_rate": 4.673090140100449e-05,
+      "loss": 1.4344,
+      "step": 12377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.021024227142334,
+      "learning_rate": 4.673063706053397e-05,
+      "loss": 1.4918,
+      "step": 12378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.620546579360962,
+      "learning_rate": 4.673037272006344e-05,
+      "loss": 1.473,
+      "step": 12379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9175488948822021,
+      "learning_rate": 4.673010837959292e-05,
+      "loss": 1.5388,
+      "step": 12380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8400986194610596,
+      "learning_rate": 4.672984403912239e-05,
+      "loss": 1.3534,
+      "step": 12381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4822447299957275,
+      "learning_rate": 4.672957969865187e-05,
+      "loss": 1.4319,
+      "step": 12382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5698466300964355,
+      "learning_rate": 4.672931535818134e-05,
+      "loss": 1.4678,
+      "step": 12383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7400765419006348,
+      "learning_rate": 4.6729051017710814e-05,
+      "loss": 1.5768,
+      "step": 12384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8430732488632202,
+      "learning_rate": 4.672878667724029e-05,
+      "loss": 1.502,
+      "step": 12385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.453169345855713,
+      "learning_rate": 4.672852233676976e-05,
+      "loss": 1.4228,
+      "step": 12386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2491493225097656,
+      "learning_rate": 4.6728257996299235e-05,
+      "loss": 1.4927,
+      "step": 12387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.71322762966156,
+      "learning_rate": 4.672799365582871e-05,
+      "loss": 1.2588,
+      "step": 12388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2615504264831543,
+      "learning_rate": 4.672772931535818e-05,
+      "loss": 1.5148,
+      "step": 12389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7589155435562134,
+      "learning_rate": 4.6727464974887655e-05,
+      "loss": 1.4993,
+      "step": 12390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3584964275360107,
+      "learning_rate": 4.6727200634417136e-05,
+      "loss": 1.1507,
+      "step": 12391
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0251262187957764,
+      "learning_rate": 4.67269362939466e-05,
+      "loss": 1.4844,
+      "step": 12392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5720359086990356,
+      "learning_rate": 4.672667195347608e-05,
+      "loss": 1.3109,
+      "step": 12393
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.866392970085144,
+      "learning_rate": 4.672640761300555e-05,
+      "loss": 1.4567,
+      "step": 12394
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.645377278327942,
+      "learning_rate": 4.672614327253503e-05,
+      "loss": 1.4851,
+      "step": 12395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5818458795547485,
+      "learning_rate": 4.6725878932064504e-05,
+      "loss": 1.3275,
+      "step": 12396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5399837493896484,
+      "learning_rate": 4.672561459159398e-05,
+      "loss": 1.2025,
+      "step": 12397
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7481728792190552,
+      "learning_rate": 4.672535025112345e-05,
+      "loss": 1.2599,
+      "step": 12398
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9831786155700684,
+      "learning_rate": 4.6725085910652924e-05,
+      "loss": 1.2689,
+      "step": 12399
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7163052558898926,
+      "learning_rate": 4.67248215701824e-05,
+      "loss": 1.4367,
+      "step": 12400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8405710458755493,
+      "learning_rate": 4.672455722971187e-05,
+      "loss": 1.4698,
+      "step": 12401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6508952379226685,
+      "learning_rate": 4.6724292889241345e-05,
+      "loss": 1.4851,
+      "step": 12402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.910101294517517,
+      "learning_rate": 4.672402854877082e-05,
+      "loss": 2.6308,
+      "step": 12403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6587886810302734,
+      "learning_rate": 4.672376420830029e-05,
+      "loss": 1.5067,
+      "step": 12404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5508687496185303,
+      "learning_rate": 4.6723499867829766e-05,
+      "loss": 1.284,
+      "step": 12405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4626483917236328,
+      "learning_rate": 4.6723235527359246e-05,
+      "loss": 1.331,
+      "step": 12406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9441221952438354,
+      "learning_rate": 4.672297118688871e-05,
+      "loss": 1.4925,
+      "step": 12407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2467329502105713,
+      "learning_rate": 4.672270684641819e-05,
+      "loss": 1.6451,
+      "step": 12408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1067111492156982,
+      "learning_rate": 4.672244250594766e-05,
+      "loss": 1.5199,
+      "step": 12409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4772493839263916,
+      "learning_rate": 4.672217816547714e-05,
+      "loss": 1.4072,
+      "step": 12410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.812771201133728,
+      "learning_rate": 4.6721913825006614e-05,
+      "loss": 1.3304,
+      "step": 12411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7606792449951172,
+      "learning_rate": 4.672164948453608e-05,
+      "loss": 1.3385,
+      "step": 12412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.520885944366455,
+      "learning_rate": 4.672138514406556e-05,
+      "loss": 1.1977,
+      "step": 12413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5590308904647827,
+      "learning_rate": 4.672112080359503e-05,
+      "loss": 1.4228,
+      "step": 12414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.612342357635498,
+      "learning_rate": 4.672085646312451e-05,
+      "loss": 1.4085,
+      "step": 12415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7190418243408203,
+      "learning_rate": 4.672059212265398e-05,
+      "loss": 1.2769,
+      "step": 12416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8774361610412598,
+      "learning_rate": 4.6720327782183455e-05,
+      "loss": 1.3136,
+      "step": 12417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5363502502441406,
+      "learning_rate": 4.672006344171293e-05,
+      "loss": 1.3545,
+      "step": 12418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7474892139434814,
+      "learning_rate": 4.67197991012424e-05,
+      "loss": 1.5173,
+      "step": 12419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.435655951499939,
+      "learning_rate": 4.6719534760771876e-05,
+      "loss": 1.3252,
+      "step": 12420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5766799449920654,
+      "learning_rate": 4.6719270420301356e-05,
+      "loss": 1.3757,
+      "step": 12421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5975291728973389,
+      "learning_rate": 4.671900607983082e-05,
+      "loss": 1.4397,
+      "step": 12422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3952369689941406,
+      "learning_rate": 4.67187417393603e-05,
+      "loss": 1.5969,
+      "step": 12423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.063481092453003,
+      "learning_rate": 4.671847739888977e-05,
+      "loss": 1.4847,
+      "step": 12424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8031303882598877,
+      "learning_rate": 4.6718213058419244e-05,
+      "loss": 1.4516,
+      "step": 12425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.195427894592285,
+      "learning_rate": 4.671794871794872e-05,
+      "loss": 1.5274,
+      "step": 12426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0273663997650146,
+      "learning_rate": 4.671768437747819e-05,
+      "loss": 1.6209,
+      "step": 12427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8236850500106812,
+      "learning_rate": 4.671742003700767e-05,
+      "loss": 1.416,
+      "step": 12428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6200412511825562,
+      "learning_rate": 4.671715569653714e-05,
+      "loss": 1.5307,
+      "step": 12429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7985447645187378,
+      "learning_rate": 4.671689135606662e-05,
+      "loss": 1.4142,
+      "step": 12430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4911279678344727,
+      "learning_rate": 4.6716627015596085e-05,
+      "loss": 1.2644,
+      "step": 12431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2269749641418457,
+      "learning_rate": 4.6716362675125566e-05,
+      "loss": 1.6133,
+      "step": 12432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7399530410766602,
+      "learning_rate": 4.671609833465504e-05,
+      "loss": 1.5395,
+      "step": 12433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.193999767303467,
+      "learning_rate": 4.671583399418451e-05,
+      "loss": 1.6273,
+      "step": 12434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8178926706314087,
+      "learning_rate": 4.6715569653713986e-05,
+      "loss": 1.5241,
+      "step": 12435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.600577712059021,
+      "learning_rate": 4.671530531324346e-05,
+      "loss": 1.2094,
+      "step": 12436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.262531042098999,
+      "learning_rate": 4.6715040972772934e-05,
+      "loss": 1.6359,
+      "step": 12437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7039873600006104,
+      "learning_rate": 4.671477663230241e-05,
+      "loss": 1.3456,
+      "step": 12438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.795086145401001,
+      "learning_rate": 4.671451229183188e-05,
+      "loss": 1.2944,
+      "step": 12439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6333332061767578,
+      "learning_rate": 4.6714247951361354e-05,
+      "loss": 1.4867,
+      "step": 12440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4588569402694702,
+      "learning_rate": 4.671398361089083e-05,
+      "loss": 1.5065,
+      "step": 12441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7450438737869263,
+      "learning_rate": 4.67137192704203e-05,
+      "loss": 1.38,
+      "step": 12442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5460655689239502,
+      "learning_rate": 4.671345492994978e-05,
+      "loss": 1.2032,
+      "step": 12443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4783730506896973,
+      "learning_rate": 4.671319058947925e-05,
+      "loss": 1.3363,
+      "step": 12444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8422794342041016,
+      "learning_rate": 4.671292624900873e-05,
+      "loss": 1.3536,
+      "step": 12445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.547252655029297,
+      "learning_rate": 4.6712661908538196e-05,
+      "loss": 1.6845,
+      "step": 12446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3542613983154297,
+      "learning_rate": 4.6712397568067676e-05,
+      "loss": 1.9182,
+      "step": 12447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2649011611938477,
+      "learning_rate": 4.671213322759715e-05,
+      "loss": 1.5525,
+      "step": 12448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6686859130859375,
+      "learning_rate": 4.6711868887126616e-05,
+      "loss": 1.2352,
+      "step": 12449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.322444438934326,
+      "learning_rate": 4.67116045466561e-05,
+      "loss": 1.4416,
+      "step": 12450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9978275299072266,
+      "learning_rate": 4.6711340206185564e-05,
+      "loss": 1.5174,
+      "step": 12451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7842878103256226,
+      "learning_rate": 4.6711075865715044e-05,
+      "loss": 1.3209,
+      "step": 12452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0609612464904785,
+      "learning_rate": 4.671081152524452e-05,
+      "loss": 1.487,
+      "step": 12453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7779746055603027,
+      "learning_rate": 4.671054718477399e-05,
+      "loss": 1.4344,
+      "step": 12454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.627078652381897,
+      "learning_rate": 4.6710282844303465e-05,
+      "loss": 1.3225,
+      "step": 12455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.445124864578247,
+      "learning_rate": 4.671001850383294e-05,
+      "loss": 1.6375,
+      "step": 12456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1514081954956055,
+      "learning_rate": 4.670975416336241e-05,
+      "loss": 1.5453,
+      "step": 12457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421522378921509,
+      "learning_rate": 4.6709489822891885e-05,
+      "loss": 1.5746,
+      "step": 12458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.240431547164917,
+      "learning_rate": 4.670922548242136e-05,
+      "loss": 1.4138,
+      "step": 12459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5914556980133057,
+      "learning_rate": 4.670896114195083e-05,
+      "loss": 1.35,
+      "step": 12460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8442143201828003,
+      "learning_rate": 4.6708696801480306e-05,
+      "loss": 1.5583,
+      "step": 12461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3316771984100342,
+      "learning_rate": 4.670843246100978e-05,
+      "loss": 1.2428,
+      "step": 12462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8659396171569824,
+      "learning_rate": 4.670816812053926e-05,
+      "loss": 1.6658,
+      "step": 12463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7191087007522583,
+      "learning_rate": 4.670790378006873e-05,
+      "loss": 1.3124,
+      "step": 12464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9642267227172852,
+      "learning_rate": 4.670763943959821e-05,
+      "loss": 1.5025,
+      "step": 12465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6543465852737427,
+      "learning_rate": 4.6707375099127674e-05,
+      "loss": 1.4298,
+      "step": 12466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7288451194763184,
+      "learning_rate": 4.6707110758657154e-05,
+      "loss": 1.6162,
+      "step": 12467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7967604398727417,
+      "learning_rate": 4.670684641818663e-05,
+      "loss": 1.5607,
+      "step": 12468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7570111751556396,
+      "learning_rate": 4.67065820777161e-05,
+      "loss": 1.4531,
+      "step": 12469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6300445795059204,
+      "learning_rate": 4.6706317737245575e-05,
+      "loss": 1.3687,
+      "step": 12470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6684386730194092,
+      "learning_rate": 4.670605339677505e-05,
+      "loss": 1.2282,
+      "step": 12471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9968587160110474,
+      "learning_rate": 4.670578905630452e-05,
+      "loss": 1.3526,
+      "step": 12472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.421067714691162,
+      "learning_rate": 4.6705524715833996e-05,
+      "loss": 1.2403,
+      "step": 12473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6266274452209473,
+      "learning_rate": 4.670526037536347e-05,
+      "loss": 1.3712,
+      "step": 12474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.469521999359131,
+      "learning_rate": 4.670499603489294e-05,
+      "loss": 1.4913,
+      "step": 12475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9021601676940918,
+      "learning_rate": 4.6704731694422416e-05,
+      "loss": 1.4683,
+      "step": 12476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0070436000823975,
+      "learning_rate": 4.670446735395189e-05,
+      "loss": 1.5354,
+      "step": 12477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0319900512695312,
+      "learning_rate": 4.670420301348137e-05,
+      "loss": 1.4117,
+      "step": 12478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1273245811462402,
+      "learning_rate": 4.670393867301084e-05,
+      "loss": 1.3285,
+      "step": 12479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5650160312652588,
+      "learning_rate": 4.670367433254032e-05,
+      "loss": 1.3101,
+      "step": 12480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1613733768463135,
+      "learning_rate": 4.6703409992069784e-05,
+      "loss": 2.0316,
+      "step": 12481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0295138359069824,
+      "learning_rate": 4.6703145651599265e-05,
+      "loss": 1.4373,
+      "step": 12482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6627684831619263,
+      "learning_rate": 4.670288131112874e-05,
+      "loss": 1.4046,
+      "step": 12483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.505769968032837,
+      "learning_rate": 4.670261697065821e-05,
+      "loss": 2.0484,
+      "step": 12484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.513972520828247,
+      "learning_rate": 4.6702352630187685e-05,
+      "loss": 1.4962,
+      "step": 12485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1297264099121094,
+      "learning_rate": 4.670208828971715e-05,
+      "loss": 1.406,
+      "step": 12486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4385007619857788,
+      "learning_rate": 4.670182394924663e-05,
+      "loss": 1.4022,
+      "step": 12487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.457927942276001,
+      "learning_rate": 4.6701559608776106e-05,
+      "loss": 1.5144,
+      "step": 12488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6836202144622803,
+      "learning_rate": 4.670129526830558e-05,
+      "loss": 1.4239,
+      "step": 12489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7939149141311646,
+      "learning_rate": 4.670103092783505e-05,
+      "loss": 1.3561,
+      "step": 12490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.995625615119934,
+      "learning_rate": 4.670076658736453e-05,
+      "loss": 1.4081,
+      "step": 12491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8452978134155273,
+      "learning_rate": 4.6700502246894e-05,
+      "loss": 1.2573,
+      "step": 12492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9529306888580322,
+      "learning_rate": 4.670023790642348e-05,
+      "loss": 1.4381,
+      "step": 12493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6991254091262817,
+      "learning_rate": 4.669997356595295e-05,
+      "loss": 1.5351,
+      "step": 12494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7110041379928589,
+      "learning_rate": 4.669970922548242e-05,
+      "loss": 1.4089,
+      "step": 12495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8574700355529785,
+      "learning_rate": 4.6699444885011895e-05,
+      "loss": 1.5269,
+      "step": 12496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6865712404251099,
+      "learning_rate": 4.669918054454137e-05,
+      "loss": 1.3223,
+      "step": 12497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8371952772140503,
+      "learning_rate": 4.669891620407085e-05,
+      "loss": 1.232,
+      "step": 12498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2851781845092773,
+      "learning_rate": 4.6698651863600315e-05,
+      "loss": 1.5257,
+      "step": 12499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9838043451309204,
+      "learning_rate": 4.6698387523129796e-05,
+      "loss": 1.5907,
+      "step": 12500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.210200548171997,
+      "learning_rate": 4.669812318265926e-05,
+      "loss": 1.3924,
+      "step": 12501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8171170949935913,
+      "learning_rate": 4.669785884218874e-05,
+      "loss": 1.406,
+      "step": 12502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6358940601348877,
+      "learning_rate": 4.6697594501718216e-05,
+      "loss": 1.376,
+      "step": 12503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3897666931152344,
+      "learning_rate": 4.669733016124769e-05,
+      "loss": 1.5312,
+      "step": 12504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9216903448104858,
+      "learning_rate": 4.6697065820777164e-05,
+      "loss": 1.3067,
+      "step": 12505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5897247791290283,
+      "learning_rate": 4.669680148030664e-05,
+      "loss": 1.3227,
+      "step": 12506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6313159465789795,
+      "learning_rate": 4.669653713983611e-05,
+      "loss": 1.4559,
+      "step": 12507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0000429153442383,
+      "learning_rate": 4.6696272799365584e-05,
+      "loss": 1.3428,
+      "step": 12508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4618114233016968,
+      "learning_rate": 4.669600845889506e-05,
+      "loss": 1.3849,
+      "step": 12509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.526831865310669,
+      "learning_rate": 4.669574411842453e-05,
+      "loss": 1.2687,
+      "step": 12510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2903261184692383,
+      "learning_rate": 4.6695479777954005e-05,
+      "loss": 1.4027,
+      "step": 12511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5205211639404297,
+      "learning_rate": 4.669521543748348e-05,
+      "loss": 1.4812,
+      "step": 12512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4455721378326416,
+      "learning_rate": 4.669495109701296e-05,
+      "loss": 1.3109,
+      "step": 12513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0487420558929443,
+      "learning_rate": 4.6694686756542426e-05,
+      "loss": 1.5276,
+      "step": 12514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.85515296459198,
+      "learning_rate": 4.6694422416071906e-05,
+      "loss": 1.456,
+      "step": 12515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3087313175201416,
+      "learning_rate": 4.669415807560137e-05,
+      "loss": 1.4615,
+      "step": 12516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.372526168823242,
+      "learning_rate": 4.669389373513085e-05,
+      "loss": 1.436,
+      "step": 12517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5331077575683594,
+      "learning_rate": 4.669362939466033e-05,
+      "loss": 1.3207,
+      "step": 12518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7178401947021484,
+      "learning_rate": 4.66933650541898e-05,
+      "loss": 1.3346,
+      "step": 12519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.75690758228302,
+      "learning_rate": 4.6693100713719274e-05,
+      "loss": 1.466,
+      "step": 12520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8058421611785889,
+      "learning_rate": 4.669283637324874e-05,
+      "loss": 1.3195,
+      "step": 12521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.696448564529419,
+      "learning_rate": 4.669257203277822e-05,
+      "loss": 1.3303,
+      "step": 12522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2031543254852295,
+      "learning_rate": 4.6692307692307695e-05,
+      "loss": 1.5728,
+      "step": 12523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.089341640472412,
+      "learning_rate": 4.669204335183717e-05,
+      "loss": 1.5646,
+      "step": 12524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6805775165557861,
+      "learning_rate": 4.669177901136664e-05,
+      "loss": 1.3709,
+      "step": 12525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8673847913742065,
+      "learning_rate": 4.6691514670896115e-05,
+      "loss": 1.4254,
+      "step": 12526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.511664867401123,
+      "learning_rate": 4.669125033042559e-05,
+      "loss": 1.3158,
+      "step": 12527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.775891900062561,
+      "learning_rate": 4.669098598995507e-05,
+      "loss": 1.321,
+      "step": 12528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6375600099563599,
+      "learning_rate": 4.6690721649484536e-05,
+      "loss": 1.3704,
+      "step": 12529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6898248195648193,
+      "learning_rate": 4.669045730901401e-05,
+      "loss": 1.3826,
+      "step": 12530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.74517822265625,
+      "learning_rate": 4.669019296854348e-05,
+      "loss": 1.5864,
+      "step": 12531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.338318347930908,
+      "learning_rate": 4.668992862807296e-05,
+      "loss": 1.4486,
+      "step": 12532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4552409648895264,
+      "learning_rate": 4.668966428760244e-05,
+      "loss": 1.5016,
+      "step": 12533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7183598279953003,
+      "learning_rate": 4.6689399947131904e-05,
+      "loss": 2.3326,
+      "step": 12534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8970810174942017,
+      "learning_rate": 4.6689135606661384e-05,
+      "loss": 1.428,
+      "step": 12535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7898563146591187,
+      "learning_rate": 4.668887126619085e-05,
+      "loss": 1.3604,
+      "step": 12536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.86042857170105,
+      "learning_rate": 4.668860692572033e-05,
+      "loss": 1.5186,
+      "step": 12537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.757656216621399,
+      "learning_rate": 4.6688342585249805e-05,
+      "loss": 1.3529,
+      "step": 12538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.739994525909424,
+      "learning_rate": 4.668807824477928e-05,
+      "loss": 1.5111,
+      "step": 12539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6702224016189575,
+      "learning_rate": 4.668781390430875e-05,
+      "loss": 1.3129,
+      "step": 12540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.681757926940918,
+      "learning_rate": 4.6687549563838226e-05,
+      "loss": 1.3158,
+      "step": 12541
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2362477779388428,
+      "learning_rate": 4.66872852233677e-05,
+      "loss": 1.5465,
+      "step": 12542
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8378069400787354,
+      "learning_rate": 4.668702088289717e-05,
+      "loss": 1.4252,
+      "step": 12543
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.969685673713684,
+      "learning_rate": 4.6686756542426646e-05,
+      "loss": 1.4707,
+      "step": 12544
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8749516010284424,
+      "learning_rate": 4.668649220195612e-05,
+      "loss": 1.3634,
+      "step": 12545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.114973306655884,
+      "learning_rate": 4.6686227861485593e-05,
+      "loss": 1.4415,
+      "step": 12546
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.576959490776062,
+      "learning_rate": 4.668596352101507e-05,
+      "loss": 1.3073,
+      "step": 12547
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4676543474197388,
+      "learning_rate": 4.668569918054455e-05,
+      "loss": 1.2964,
+      "step": 12548
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6518850326538086,
+      "learning_rate": 4.6685434840074014e-05,
+      "loss": 1.4937,
+      "step": 12549
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9786860942840576,
+      "learning_rate": 4.6685170499603495e-05,
+      "loss": 1.403,
+      "step": 12550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.643606424331665,
+      "learning_rate": 4.668490615913296e-05,
+      "loss": 1.3365,
+      "step": 12551
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.940246820449829,
+      "learning_rate": 4.668464181866244e-05,
+      "loss": 1.4279,
+      "step": 12552
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.380049228668213,
+      "learning_rate": 4.6684377478191915e-05,
+      "loss": 1.3731,
+      "step": 12553
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.153839588165283,
+      "learning_rate": 4.668411313772139e-05,
+      "loss": 1.4695,
+      "step": 12554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8497858047485352,
+      "learning_rate": 4.668384879725086e-05,
+      "loss": 1.5158,
+      "step": 12555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5748363733291626,
+      "learning_rate": 4.6683584456780336e-05,
+      "loss": 1.4767,
+      "step": 12556
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8854899406433105,
+      "learning_rate": 4.668332011630981e-05,
+      "loss": 1.514,
+      "step": 12557
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6329141855239868,
+      "learning_rate": 4.668305577583928e-05,
+      "loss": 1.3107,
+      "step": 12558
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.804585337638855,
+      "learning_rate": 4.668279143536876e-05,
+      "loss": 1.3202,
+      "step": 12559
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7100614309310913,
+      "learning_rate": 4.668252709489823e-05,
+      "loss": 1.2434,
+      "step": 12560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3186206817626953,
+      "learning_rate": 4.6682262754427704e-05,
+      "loss": 1.4129,
+      "step": 12561
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4072072505950928,
+      "learning_rate": 4.668199841395718e-05,
+      "loss": 1.5407,
+      "step": 12562
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1126227378845215,
+      "learning_rate": 4.668173407348666e-05,
+      "loss": 1.4994,
+      "step": 12563
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.027158737182617,
+      "learning_rate": 4.6681469733016125e-05,
+      "loss": 1.5296,
+      "step": 12564
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.680364966392517,
+      "learning_rate": 4.6681205392545605e-05,
+      "loss": 1.2532,
+      "step": 12565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4369583129882812,
+      "learning_rate": 4.668094105207507e-05,
+      "loss": 1.4483,
+      "step": 12566
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.560902714729309,
+      "learning_rate": 4.6680676711604545e-05,
+      "loss": 1.3497,
+      "step": 12567
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1983559131622314,
+      "learning_rate": 4.6680412371134026e-05,
+      "loss": 1.3562,
+      "step": 12568
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8652727603912354,
+      "learning_rate": 4.668014803066349e-05,
+      "loss": 1.2015,
+      "step": 12569
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.48498797416687,
+      "learning_rate": 4.667988369019297e-05,
+      "loss": 1.3578,
+      "step": 12570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2839155197143555,
+      "learning_rate": 4.667961934972244e-05,
+      "loss": 1.7362,
+      "step": 12571
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7859148979187012,
+      "learning_rate": 4.667935500925192e-05,
+      "loss": 1.4486,
+      "step": 12572
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6663413047790527,
+      "learning_rate": 4.6679090668781393e-05,
+      "loss": 1.099,
+      "step": 12573
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9431071281433105,
+      "learning_rate": 4.667882632831087e-05,
+      "loss": 1.3333,
+      "step": 12574
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9102016687393188,
+      "learning_rate": 4.667856198784034e-05,
+      "loss": 1.3197,
+      "step": 12575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6428544521331787,
+      "learning_rate": 4.6678297647369814e-05,
+      "loss": 1.4547,
+      "step": 12576
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7664339542388916,
+      "learning_rate": 4.667803330689929e-05,
+      "loss": 1.4397,
+      "step": 12577
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.238659143447876,
+      "learning_rate": 4.667776896642876e-05,
+      "loss": 1.5159,
+      "step": 12578
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.149139165878296,
+      "learning_rate": 4.6677504625958235e-05,
+      "loss": 1.4588,
+      "step": 12579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6963069438934326,
+      "learning_rate": 4.667724028548771e-05,
+      "loss": 1.4769,
+      "step": 12580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0565543174743652,
+      "learning_rate": 4.667697594501718e-05,
+      "loss": 1.4608,
+      "step": 12581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2602086067199707,
+      "learning_rate": 4.6676711604546656e-05,
+      "loss": 1.3739,
+      "step": 12582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7994611263275146,
+      "learning_rate": 4.6676447264076136e-05,
+      "loss": 1.505,
+      "step": 12583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8436291217803955,
+      "learning_rate": 4.66761829236056e-05,
+      "loss": 1.514,
+      "step": 12584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9348965883255005,
+      "learning_rate": 4.667591858313508e-05,
+      "loss": 1.5022,
+      "step": 12585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21262526512146,
+      "learning_rate": 4.667565424266455e-05,
+      "loss": 1.3969,
+      "step": 12586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.565443754196167,
+      "learning_rate": 4.667538990219403e-05,
+      "loss": 1.4231,
+      "step": 12587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.118783712387085,
+      "learning_rate": 4.6675125561723504e-05,
+      "loss": 1.4361,
+      "step": 12588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.205376148223877,
+      "learning_rate": 4.667486122125298e-05,
+      "loss": 1.4983,
+      "step": 12589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0297446250915527,
+      "learning_rate": 4.667459688078245e-05,
+      "loss": 1.7346,
+      "step": 12590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6954476833343506,
+      "learning_rate": 4.6674332540311925e-05,
+      "loss": 1.3942,
+      "step": 12591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6640989780426025,
+      "learning_rate": 4.66740681998414e-05,
+      "loss": 1.3638,
+      "step": 12592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.659438133239746,
+      "learning_rate": 4.667380385937087e-05,
+      "loss": 1.3712,
+      "step": 12593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.040355682373047,
+      "learning_rate": 4.6673539518900345e-05,
+      "loss": 1.3698,
+      "step": 12594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9835032224655151,
+      "learning_rate": 4.667327517842982e-05,
+      "loss": 1.4376,
+      "step": 12595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4731873273849487,
+      "learning_rate": 4.667301083795929e-05,
+      "loss": 1.4065,
+      "step": 12596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4745349884033203,
+      "learning_rate": 4.6672746497488766e-05,
+      "loss": 1.1701,
+      "step": 12597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.266096830368042,
+      "learning_rate": 4.6672482157018246e-05,
+      "loss": 1.3584,
+      "step": 12598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7106391191482544,
+      "learning_rate": 4.667221781654771e-05,
+      "loss": 1.397,
+      "step": 12599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5241310596466064,
+      "learning_rate": 4.6671953476077193e-05,
+      "loss": 1.3699,
+      "step": 12600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8977311849594116,
+      "learning_rate": 4.667168913560666e-05,
+      "loss": 1.3553,
+      "step": 12601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4109363555908203,
+      "learning_rate": 4.667142479513614e-05,
+      "loss": 1.3389,
+      "step": 12602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.646641731262207,
+      "learning_rate": 4.6671160454665614e-05,
+      "loss": 1.6121,
+      "step": 12603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5773849487304688,
+      "learning_rate": 4.667089611419508e-05,
+      "loss": 1.4831,
+      "step": 12604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6063100099563599,
+      "learning_rate": 4.667063177372456e-05,
+      "loss": 1.4066,
+      "step": 12605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5912684202194214,
+      "learning_rate": 4.667036743325403e-05,
+      "loss": 1.4018,
+      "step": 12606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.767978310585022,
+      "learning_rate": 4.667010309278351e-05,
+      "loss": 1.3857,
+      "step": 12607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6079063415527344,
+      "learning_rate": 4.666983875231298e-05,
+      "loss": 1.4868,
+      "step": 12608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9149330854415894,
+      "learning_rate": 4.6669574411842456e-05,
+      "loss": 1.7285,
+      "step": 12609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.048645496368408,
+      "learning_rate": 4.666931007137193e-05,
+      "loss": 1.5334,
+      "step": 12610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0178463459014893,
+      "learning_rate": 4.66690457309014e-05,
+      "loss": 1.4194,
+      "step": 12611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6977907419204712,
+      "learning_rate": 4.6668781390430876e-05,
+      "loss": 1.4459,
+      "step": 12612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8997881412506104,
+      "learning_rate": 4.666851704996035e-05,
+      "loss": 1.2305,
+      "step": 12613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5497621297836304,
+      "learning_rate": 4.6668252709489823e-05,
+      "loss": 1.3078,
+      "step": 12614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1981379985809326,
+      "learning_rate": 4.66679883690193e-05,
+      "loss": 1.4589,
+      "step": 12615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7852377891540527,
+      "learning_rate": 4.666772402854877e-05,
+      "loss": 1.1767,
+      "step": 12616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7427006959915161,
+      "learning_rate": 4.6667459688078244e-05,
+      "loss": 2.0107,
+      "step": 12617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2884936332702637,
+      "learning_rate": 4.6667195347607724e-05,
+      "loss": 1.4826,
+      "step": 12618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.936216115951538,
+      "learning_rate": 4.666693100713719e-05,
+      "loss": 1.3046,
+      "step": 12619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7994930744171143,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 1.4156,
+      "step": 12620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6895164251327515,
+      "learning_rate": 4.666640232619614e-05,
+      "loss": 1.7315,
+      "step": 12621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6759004592895508,
+      "learning_rate": 4.666613798572562e-05,
+      "loss": 1.3834,
+      "step": 12622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1468942165374756,
+      "learning_rate": 4.666587364525509e-05,
+      "loss": 1.5781,
+      "step": 12623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8582422733306885,
+      "learning_rate": 4.6665609304784566e-05,
+      "loss": 1.422,
+      "step": 12624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9677244424819946,
+      "learning_rate": 4.666534496431404e-05,
+      "loss": 1.2941,
+      "step": 12625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.585145950317383,
+      "learning_rate": 4.666508062384351e-05,
+      "loss": 1.4941,
+      "step": 12626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4850645065307617,
+      "learning_rate": 4.6664816283372987e-05,
+      "loss": 1.3499,
+      "step": 12627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7821837663650513,
+      "learning_rate": 4.666455194290246e-05,
+      "loss": 1.3918,
+      "step": 12628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.711912989616394,
+      "learning_rate": 4.6664287602431934e-05,
+      "loss": 1.3888,
+      "step": 12629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9387431144714355,
+      "learning_rate": 4.666402326196141e-05,
+      "loss": 1.3846,
+      "step": 12630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.944477915763855,
+      "learning_rate": 4.666375892149088e-05,
+      "loss": 1.4485,
+      "step": 12631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6444047689437866,
+      "learning_rate": 4.6663494581020354e-05,
+      "loss": 1.4027,
+      "step": 12632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7946223020553589,
+      "learning_rate": 4.6663230240549835e-05,
+      "loss": 1.3974,
+      "step": 12633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7342438697814941,
+      "learning_rate": 4.66629659000793e-05,
+      "loss": 1.3127,
+      "step": 12634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9321684837341309,
+      "learning_rate": 4.666270155960878e-05,
+      "loss": 1.5212,
+      "step": 12635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.73298978805542,
+      "learning_rate": 4.666243721913825e-05,
+      "loss": 1.4046,
+      "step": 12636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4724980592727661,
+      "learning_rate": 4.666217287866773e-05,
+      "loss": 1.2303,
+      "step": 12637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4731590747833252,
+      "learning_rate": 4.66619085381972e-05,
+      "loss": 1.2263,
+      "step": 12638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.451512098312378,
+      "learning_rate": 4.666164419772667e-05,
+      "loss": 1.4198,
+      "step": 12639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4208879470825195,
+      "learning_rate": 4.666137985725615e-05,
+      "loss": 1.4665,
+      "step": 12640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7695066928863525,
+      "learning_rate": 4.6661115516785617e-05,
+      "loss": 1.3182,
+      "step": 12641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.919647455215454,
+      "learning_rate": 4.66608511763151e-05,
+      "loss": 2.0669,
+      "step": 12642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9509472846984863,
+      "learning_rate": 4.666058683584457e-05,
+      "loss": 1.3922,
+      "step": 12643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.037613868713379,
+      "learning_rate": 4.6660322495374044e-05,
+      "loss": 1.3975,
+      "step": 12644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7446073293685913,
+      "learning_rate": 4.666005815490352e-05,
+      "loss": 1.4482,
+      "step": 12645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.690618634223938,
+      "learning_rate": 4.665979381443299e-05,
+      "loss": 1.3918,
+      "step": 12646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9041897058486938,
+      "learning_rate": 4.6659529473962465e-05,
+      "loss": 1.4445,
+      "step": 12647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9764909744262695,
+      "learning_rate": 4.665926513349194e-05,
+      "loss": 1.2972,
+      "step": 12648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4573475122451782,
+      "learning_rate": 4.665900079302141e-05,
+      "loss": 1.3689,
+      "step": 12649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.531690239906311,
+      "learning_rate": 4.6658736452550886e-05,
+      "loss": 1.1916,
+      "step": 12650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.668648600578308,
+      "learning_rate": 4.665847211208036e-05,
+      "loss": 1.425,
+      "step": 12651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4883759021759033,
+      "learning_rate": 4.665820777160983e-05,
+      "loss": 1.4742,
+      "step": 12652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9641162157058716,
+      "learning_rate": 4.665794343113931e-05,
+      "loss": 1.4954,
+      "step": 12653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0792794227600098,
+      "learning_rate": 4.665767909066878e-05,
+      "loss": 1.4832,
+      "step": 12654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3336418867111206,
+      "learning_rate": 4.665741475019826e-05,
+      "loss": 1.4378,
+      "step": 12655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6464262008666992,
+      "learning_rate": 4.665715040972773e-05,
+      "loss": 1.3986,
+      "step": 12656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6134682893753052,
+      "learning_rate": 4.665688606925721e-05,
+      "loss": 1.4162,
+      "step": 12657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5837825536727905,
+      "learning_rate": 4.665662172878668e-05,
+      "loss": 1.3762,
+      "step": 12658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9685134887695312,
+      "learning_rate": 4.6656357388316154e-05,
+      "loss": 1.2815,
+      "step": 12659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4302157163619995,
+      "learning_rate": 4.665609304784563e-05,
+      "loss": 2.0026,
+      "step": 12660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6447985172271729,
+      "learning_rate": 4.66558287073751e-05,
+      "loss": 1.5025,
+      "step": 12661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.914778232574463,
+      "learning_rate": 4.6655564366904575e-05,
+      "loss": 1.5532,
+      "step": 12662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2872892618179321,
+      "learning_rate": 4.665530002643405e-05,
+      "loss": 1.2008,
+      "step": 12663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5764025449752808,
+      "learning_rate": 4.665503568596352e-05,
+      "loss": 1.3698,
+      "step": 12664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9963715076446533,
+      "learning_rate": 4.6654771345492996e-05,
+      "loss": 1.3362,
+      "step": 12665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5837482213974,
+      "learning_rate": 4.665450700502247e-05,
+      "loss": 1.3396,
+      "step": 12666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9987494945526123,
+      "learning_rate": 4.665424266455194e-05,
+      "loss": 1.4485,
+      "step": 12667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8857192993164062,
+      "learning_rate": 4.665397832408142e-05,
+      "loss": 1.559,
+      "step": 12668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8086522817611694,
+      "learning_rate": 4.665371398361089e-05,
+      "loss": 1.3987,
+      "step": 12669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7996644973754883,
+      "learning_rate": 4.665344964314037e-05,
+      "loss": 1.4195,
+      "step": 12670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.330312967300415,
+      "learning_rate": 4.665318530266984e-05,
+      "loss": 1.5072,
+      "step": 12671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.955316424369812,
+      "learning_rate": 4.665292096219932e-05,
+      "loss": 1.3553,
+      "step": 12672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6901092529296875,
+      "learning_rate": 4.665265662172879e-05,
+      "loss": 1.2799,
+      "step": 12673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7184884548187256,
+      "learning_rate": 4.6652392281258265e-05,
+      "loss": 1.3983,
+      "step": 12674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1509556770324707,
+      "learning_rate": 4.665212794078774e-05,
+      "loss": 1.4856,
+      "step": 12675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5787020921707153,
+      "learning_rate": 4.6651863600317205e-05,
+      "loss": 1.4141,
+      "step": 12676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.623041868209839,
+      "learning_rate": 4.6651599259846685e-05,
+      "loss": 1.4649,
+      "step": 12677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.184861660003662,
+      "learning_rate": 4.665133491937616e-05,
+      "loss": 1.4106,
+      "step": 12678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4770594835281372,
+      "learning_rate": 4.665107057890563e-05,
+      "loss": 1.2751,
+      "step": 12679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6066406965255737,
+      "learning_rate": 4.6650806238435106e-05,
+      "loss": 1.3721,
+      "step": 12680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7315616607666016,
+      "learning_rate": 4.665054189796458e-05,
+      "loss": 1.2735,
+      "step": 12681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.719133734703064,
+      "learning_rate": 4.665027755749405e-05,
+      "loss": 1.3712,
+      "step": 12682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8737359046936035,
+      "learning_rate": 4.6650013217023534e-05,
+      "loss": 1.3021,
+      "step": 12683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6852948665618896,
+      "learning_rate": 4.6649748876553e-05,
+      "loss": 1.2328,
+      "step": 12684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.725036382675171,
+      "learning_rate": 4.6649484536082474e-05,
+      "loss": 1.1933,
+      "step": 12685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5502183437347412,
+      "learning_rate": 4.664922019561195e-05,
+      "loss": 1.2683,
+      "step": 12686
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3510291576385498,
+      "learning_rate": 4.664895585514142e-05,
+      "loss": 1.2465,
+      "step": 12687
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.952515721321106,
+      "learning_rate": 4.66486915146709e-05,
+      "loss": 1.46,
+      "step": 12688
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.805308222770691,
+      "learning_rate": 4.664842717420037e-05,
+      "loss": 2.318,
+      "step": 12689
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.68366277217865,
+      "learning_rate": 4.664816283372985e-05,
+      "loss": 1.4071,
+      "step": 12690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8839423656463623,
+      "learning_rate": 4.6647898493259315e-05,
+      "loss": 1.5006,
+      "step": 12691
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9247567653656006,
+      "learning_rate": 4.6647634152788796e-05,
+      "loss": 1.5062,
+      "step": 12692
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.960307002067566,
+      "learning_rate": 4.664736981231827e-05,
+      "loss": 1.5053,
+      "step": 12693
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.460007667541504,
+      "learning_rate": 4.664710547184774e-05,
+      "loss": 1.4842,
+      "step": 12694
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5521092414855957,
+      "learning_rate": 4.6646841131377217e-05,
+      "loss": 1.5257,
+      "step": 12695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2043612003326416,
+      "learning_rate": 4.664657679090669e-05,
+      "loss": 1.531,
+      "step": 12696
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.139214038848877,
+      "learning_rate": 4.6646312450436164e-05,
+      "loss": 2.4488,
+      "step": 12697
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8357139825820923,
+      "learning_rate": 4.664604810996564e-05,
+      "loss": 1.3929,
+      "step": 12698
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.402452230453491,
+      "learning_rate": 4.664578376949511e-05,
+      "loss": 1.53,
+      "step": 12699
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.948427677154541,
+      "learning_rate": 4.6645519429024584e-05,
+      "loss": 1.5227,
+      "step": 12700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2582778930664062,
+      "learning_rate": 4.664525508855406e-05,
+      "loss": 1.5277,
+      "step": 12701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1134305000305176,
+      "learning_rate": 4.664499074808353e-05,
+      "loss": 1.5311,
+      "step": 12702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6153334379196167,
+      "learning_rate": 4.664472640761301e-05,
+      "loss": 1.2764,
+      "step": 12703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6548163890838623,
+      "learning_rate": 4.664446206714248e-05,
+      "loss": 1.4348,
+      "step": 12704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0421981811523438,
+      "learning_rate": 4.664419772667196e-05,
+      "loss": 1.3815,
+      "step": 12705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8685150146484375,
+      "learning_rate": 4.6643933386201426e-05,
+      "loss": 1.42,
+      "step": 12706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1650991439819336,
+      "learning_rate": 4.6643669045730906e-05,
+      "loss": 1.471,
+      "step": 12707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9529767036437988,
+      "learning_rate": 4.664340470526038e-05,
+      "loss": 1.4654,
+      "step": 12708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.052764415740967,
+      "learning_rate": 4.664314036478985e-05,
+      "loss": 1.3716,
+      "step": 12709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.688746690750122,
+      "learning_rate": 4.664287602431933e-05,
+      "loss": 1.5546,
+      "step": 12710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3215597867965698,
+      "learning_rate": 4.6642611683848794e-05,
+      "loss": 1.2009,
+      "step": 12711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8900415897369385,
+      "learning_rate": 4.6642347343378274e-05,
+      "loss": 1.4384,
+      "step": 12712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.437278389930725,
+      "learning_rate": 4.664208300290775e-05,
+      "loss": 1.3645,
+      "step": 12713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6141210794448853,
+      "learning_rate": 4.664181866243722e-05,
+      "loss": 1.3144,
+      "step": 12714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1739048957824707,
+      "learning_rate": 4.6641554321966695e-05,
+      "loss": 1.5585,
+      "step": 12715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4270384311676025,
+      "learning_rate": 4.664128998149617e-05,
+      "loss": 1.2959,
+      "step": 12716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2616865634918213,
+      "learning_rate": 4.664102564102564e-05,
+      "loss": 1.5106,
+      "step": 12717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6367729902267456,
+      "learning_rate": 4.664076130055512e-05,
+      "loss": 1.353,
+      "step": 12718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8848434686660767,
+      "learning_rate": 4.664049696008459e-05,
+      "loss": 1.4007,
+      "step": 12719
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4933409690856934,
+      "learning_rate": 4.664023261961407e-05,
+      "loss": 1.3304,
+      "step": 12720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7497389316558838,
+      "learning_rate": 4.6639968279143536e-05,
+      "loss": 1.3393,
+      "step": 12721
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8044683933258057,
+      "learning_rate": 4.663970393867301e-05,
+      "loss": 1.4313,
+      "step": 12722
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7021547555923462,
+      "learning_rate": 4.663943959820249e-05,
+      "loss": 1.385,
+      "step": 12723
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.637585997581482,
+      "learning_rate": 4.663917525773196e-05,
+      "loss": 1.332,
+      "step": 12724
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.619688868522644,
+      "learning_rate": 4.663891091726144e-05,
+      "loss": 1.3108,
+      "step": 12725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8694862127304077,
+      "learning_rate": 4.6638646576790904e-05,
+      "loss": 1.4277,
+      "step": 12726
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8309669494628906,
+      "learning_rate": 4.6638382236320384e-05,
+      "loss": 1.3537,
+      "step": 12727
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0604469776153564,
+      "learning_rate": 4.663811789584986e-05,
+      "loss": 1.3973,
+      "step": 12728
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7186357975006104,
+      "learning_rate": 4.663785355537933e-05,
+      "loss": 1.3486,
+      "step": 12729
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5769591331481934,
+      "learning_rate": 4.6637589214908805e-05,
+      "loss": 1.2867,
+      "step": 12730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.593443512916565,
+      "learning_rate": 4.663732487443828e-05,
+      "loss": 1.3225,
+      "step": 12731
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8828791379928589,
+      "learning_rate": 4.663706053396775e-05,
+      "loss": 1.3859,
+      "step": 12732
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8521360158920288,
+      "learning_rate": 4.6636796193497226e-05,
+      "loss": 1.3677,
+      "step": 12733
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.785850763320923,
+      "learning_rate": 4.66365318530267e-05,
+      "loss": 1.5972,
+      "step": 12734
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7054396867752075,
+      "learning_rate": 4.663626751255617e-05,
+      "loss": 1.2821,
+      "step": 12735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.193734884262085,
+      "learning_rate": 4.6636003172085646e-05,
+      "loss": 1.5577,
+      "step": 12736
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.383213758468628,
+      "learning_rate": 4.663573883161512e-05,
+      "loss": 1.441,
+      "step": 12737
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9434653520584106,
+      "learning_rate": 4.66354744911446e-05,
+      "loss": 1.4872,
+      "step": 12738
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8344975709915161,
+      "learning_rate": 4.663521015067407e-05,
+      "loss": 1.3044,
+      "step": 12739
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.52252995967865,
+      "learning_rate": 4.663494581020355e-05,
+      "loss": 1.4118,
+      "step": 12740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4183471202850342,
+      "learning_rate": 4.6634681469733014e-05,
+      "loss": 1.3061,
+      "step": 12741
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5098074674606323,
+      "learning_rate": 4.6634417129262495e-05,
+      "loss": 1.3743,
+      "step": 12742
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.979218602180481,
+      "learning_rate": 4.663415278879197e-05,
+      "loss": 1.4413,
+      "step": 12743
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.063281774520874,
+      "learning_rate": 4.663388844832144e-05,
+      "loss": 1.4348,
+      "step": 12744
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5845752954483032,
+      "learning_rate": 4.6633624107850915e-05,
+      "loss": 2.2106,
+      "step": 12745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8915903568267822,
+      "learning_rate": 4.663335976738039e-05,
+      "loss": 1.3014,
+      "step": 12746
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8824403285980225,
+      "learning_rate": 4.663309542690986e-05,
+      "loss": 1.3227,
+      "step": 12747
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.640636920928955,
+      "learning_rate": 4.6632831086439336e-05,
+      "loss": 1.3322,
+      "step": 12748
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.422712802886963,
+      "learning_rate": 4.663256674596881e-05,
+      "loss": 1.2671,
+      "step": 12749
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5717259645462036,
+      "learning_rate": 4.663230240549828e-05,
+      "loss": 1.3619,
+      "step": 12750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0278379917144775,
+      "learning_rate": 4.663203806502776e-05,
+      "loss": 1.3283,
+      "step": 12751
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3302903175354004,
+      "learning_rate": 4.663177372455723e-05,
+      "loss": 1.5043,
+      "step": 12752
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8047879934310913,
+      "learning_rate": 4.663150938408671e-05,
+      "loss": 1.5237,
+      "step": 12753
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6347075700759888,
+      "learning_rate": 4.663124504361618e-05,
+      "loss": 1.2061,
+      "step": 12754
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.561751127243042,
+      "learning_rate": 4.663098070314566e-05,
+      "loss": 1.2998,
+      "step": 12755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5862387418746948,
+      "learning_rate": 4.6630716362675125e-05,
+      "loss": 1.4575,
+      "step": 12756
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.918327808380127,
+      "learning_rate": 4.66304520222046e-05,
+      "loss": 1.4126,
+      "step": 12757
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4057530164718628,
+      "learning_rate": 4.663018768173408e-05,
+      "loss": 1.2023,
+      "step": 12758
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2241153717041016,
+      "learning_rate": 4.6629923341263545e-05,
+      "loss": 1.5021,
+      "step": 12759
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9764031171798706,
+      "learning_rate": 4.6629659000793026e-05,
+      "loss": 1.5279,
+      "step": 12760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6135292053222656,
+      "learning_rate": 4.662939466032249e-05,
+      "loss": 1.3505,
+      "step": 12761
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1134369373321533,
+      "learning_rate": 4.662913031985197e-05,
+      "loss": 1.3063,
+      "step": 12762
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5739914178848267,
+      "learning_rate": 4.6628865979381446e-05,
+      "loss": 1.3025,
+      "step": 12763
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.146461248397827,
+      "learning_rate": 4.662860163891092e-05,
+      "loss": 1.5118,
+      "step": 12764
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.594309687614441,
+      "learning_rate": 4.6628337298440394e-05,
+      "loss": 1.33,
+      "step": 12765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5124280452728271,
+      "learning_rate": 4.662807295796987e-05,
+      "loss": 1.3233,
+      "step": 12766
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.439704179763794,
+      "learning_rate": 4.662780861749934e-05,
+      "loss": 1.2978,
+      "step": 12767
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.040682077407837,
+      "learning_rate": 4.6627544277028814e-05,
+      "loss": 1.4062,
+      "step": 12768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7713807821273804,
+      "learning_rate": 4.662727993655829e-05,
+      "loss": 1.3952,
+      "step": 12769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8655201196670532,
+      "learning_rate": 4.662701559608776e-05,
+      "loss": 1.4399,
+      "step": 12770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7126107215881348,
+      "learning_rate": 4.6626751255617235e-05,
+      "loss": 1.4528,
+      "step": 12771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1949729919433594,
+      "learning_rate": 4.662648691514671e-05,
+      "loss": 1.5036,
+      "step": 12772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7473427057266235,
+      "learning_rate": 4.662622257467619e-05,
+      "loss": 1.4837,
+      "step": 12773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.583304524421692,
+      "learning_rate": 4.6625958234205656e-05,
+      "loss": 1.2917,
+      "step": 12774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.224818706512451,
+      "learning_rate": 4.6625693893735136e-05,
+      "loss": 1.4559,
+      "step": 12775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.057724714279175,
+      "learning_rate": 4.66254295532646e-05,
+      "loss": 1.2502,
+      "step": 12776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6841483116149902,
+      "learning_rate": 4.662516521279408e-05,
+      "loss": 1.2257,
+      "step": 12777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1616973876953125,
+      "learning_rate": 4.662490087232356e-05,
+      "loss": 1.4383,
+      "step": 12778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9477784633636475,
+      "learning_rate": 4.662463653185303e-05,
+      "loss": 1.4966,
+      "step": 12779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6716357469558716,
+      "learning_rate": 4.6624372191382504e-05,
+      "loss": 1.4225,
+      "step": 12780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.628978967666626,
+      "learning_rate": 4.662410785091198e-05,
+      "loss": 1.4645,
+      "step": 12781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1753487586975098,
+      "learning_rate": 4.662384351044145e-05,
+      "loss": 1.5367,
+      "step": 12782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8714722394943237,
+      "learning_rate": 4.6623579169970925e-05,
+      "loss": 1.4089,
+      "step": 12783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.121119499206543,
+      "learning_rate": 4.66233148295004e-05,
+      "loss": 1.465,
+      "step": 12784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3610072135925293,
+      "learning_rate": 4.662305048902987e-05,
+      "loss": 1.3265,
+      "step": 12785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9356650114059448,
+      "learning_rate": 4.6622786148559345e-05,
+      "loss": 1.6622,
+      "step": 12786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.40139639377594,
+      "learning_rate": 4.662252180808882e-05,
+      "loss": 1.3417,
+      "step": 12787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.681607484817505,
+      "learning_rate": 4.66222574676183e-05,
+      "loss": 1.6221,
+      "step": 12788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9406076669692993,
+      "learning_rate": 4.6621993127147766e-05,
+      "loss": 1.4703,
+      "step": 12789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7304631471633911,
+      "learning_rate": 4.6621728786677246e-05,
+      "loss": 1.53,
+      "step": 12790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3710031509399414,
+      "learning_rate": 4.662146444620671e-05,
+      "loss": 1.3651,
+      "step": 12791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6758856773376465,
+      "learning_rate": 4.6621200105736194e-05,
+      "loss": 1.1726,
+      "step": 12792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7690154314041138,
+      "learning_rate": 4.662093576526567e-05,
+      "loss": 1.4757,
+      "step": 12793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.247235655784607,
+      "learning_rate": 4.6620671424795134e-05,
+      "loss": 1.1798,
+      "step": 12794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6280417442321777,
+      "learning_rate": 4.6620407084324614e-05,
+      "loss": 1.4609,
+      "step": 12795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8677533864974976,
+      "learning_rate": 4.662014274385408e-05,
+      "loss": 1.2881,
+      "step": 12796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9970048666000366,
+      "learning_rate": 4.661987840338356e-05,
+      "loss": 1.5304,
+      "step": 12797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9873169660568237,
+      "learning_rate": 4.6619614062913035e-05,
+      "loss": 1.4871,
+      "step": 12798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1039109230041504,
+      "learning_rate": 4.661934972244251e-05,
+      "loss": 1.6702,
+      "step": 12799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.452505588531494,
+      "learning_rate": 4.661908538197198e-05,
+      "loss": 1.4277,
+      "step": 12800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.838167428970337,
+      "learning_rate": 4.6618821041501456e-05,
+      "loss": 1.328,
+      "step": 12801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.296500563621521,
+      "learning_rate": 4.661855670103093e-05,
+      "loss": 1.3144,
+      "step": 12802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.62588632106781,
+      "learning_rate": 4.66182923605604e-05,
+      "loss": 1.4213,
+      "step": 12803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9267971515655518,
+      "learning_rate": 4.6618028020089876e-05,
+      "loss": 1.5165,
+      "step": 12804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6263201236724854,
+      "learning_rate": 4.661776367961935e-05,
+      "loss": 1.4129,
+      "step": 12805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4175620079040527,
+      "learning_rate": 4.6617499339148824e-05,
+      "loss": 1.2884,
+      "step": 12806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.729792594909668,
+      "learning_rate": 4.66172349986783e-05,
+      "loss": 1.3837,
+      "step": 12807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.945749044418335,
+      "learning_rate": 4.661697065820778e-05,
+      "loss": 2.4916,
+      "step": 12808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7979731559753418,
+      "learning_rate": 4.6616706317737244e-05,
+      "loss": 1.3696,
+      "step": 12809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.87261962890625,
+      "learning_rate": 4.6616441977266725e-05,
+      "loss": 1.4842,
+      "step": 12810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3420997858047485,
+      "learning_rate": 4.661617763679619e-05,
+      "loss": 1.3157,
+      "step": 12811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7063109874725342,
+      "learning_rate": 4.661591329632567e-05,
+      "loss": 1.3001,
+      "step": 12812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9580079317092896,
+      "learning_rate": 4.6615648955855145e-05,
+      "loss": 1.3288,
+      "step": 12813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4969232082366943,
+      "learning_rate": 4.661538461538462e-05,
+      "loss": 1.2936,
+      "step": 12814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8946006298065186,
+      "learning_rate": 4.661512027491409e-05,
+      "loss": 1.4653,
+      "step": 12815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.23298716545105,
+      "learning_rate": 4.6614855934443566e-05,
+      "loss": 1.5079,
+      "step": 12816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4940820932388306,
+      "learning_rate": 4.661459159397304e-05,
+      "loss": 1.2934,
+      "step": 12817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2249159812927246,
+      "learning_rate": 4.661432725350251e-05,
+      "loss": 1.3282,
+      "step": 12818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6245309114456177,
+      "learning_rate": 4.661406291303199e-05,
+      "loss": 1.3559,
+      "step": 12819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6207107305526733,
+      "learning_rate": 4.661379857256146e-05,
+      "loss": 1.4803,
+      "step": 12820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8968446254730225,
+      "learning_rate": 4.6613534232090934e-05,
+      "loss": 1.4422,
+      "step": 12821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.990427017211914,
+      "learning_rate": 4.661326989162041e-05,
+      "loss": 1.5478,
+      "step": 12822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.892419457435608,
+      "learning_rate": 4.661300555114989e-05,
+      "loss": 1.4256,
+      "step": 12823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5198235511779785,
+      "learning_rate": 4.6612741210679355e-05,
+      "loss": 1.523,
+      "step": 12824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9171359539031982,
+      "learning_rate": 4.6612476870208835e-05,
+      "loss": 1.2877,
+      "step": 12825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8746227025985718,
+      "learning_rate": 4.66122125297383e-05,
+      "loss": 1.3556,
+      "step": 12826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.681799054145813,
+      "learning_rate": 4.661194818926778e-05,
+      "loss": 1.4085,
+      "step": 12827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.104849338531494,
+      "learning_rate": 4.6611683848797256e-05,
+      "loss": 1.4221,
+      "step": 12828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5528637170791626,
+      "learning_rate": 4.661141950832672e-05,
+      "loss": 1.4526,
+      "step": 12829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1405060291290283,
+      "learning_rate": 4.66111551678562e-05,
+      "loss": 1.5315,
+      "step": 12830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6543796062469482,
+      "learning_rate": 4.661089082738567e-05,
+      "loss": 1.4812,
+      "step": 12831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.641004204750061,
+      "learning_rate": 4.661062648691515e-05,
+      "loss": 1.2415,
+      "step": 12832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8339011669158936,
+      "learning_rate": 4.6610362146444624e-05,
+      "loss": 1.3422,
+      "step": 12833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1580448150634766,
+      "learning_rate": 4.66100978059741e-05,
+      "loss": 1.4486,
+      "step": 12834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0184643268585205,
+      "learning_rate": 4.660983346550357e-05,
+      "loss": 1.3189,
+      "step": 12835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.548858880996704,
+      "learning_rate": 4.6609569125033044e-05,
+      "loss": 1.3448,
+      "step": 12836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2082958221435547,
+      "learning_rate": 4.660930478456252e-05,
+      "loss": 1.3348,
+      "step": 12837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5065916776657104,
+      "learning_rate": 4.660904044409199e-05,
+      "loss": 1.3347,
+      "step": 12838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2161507606506348,
+      "learning_rate": 4.6608776103621465e-05,
+      "loss": 1.4749,
+      "step": 12839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6284286975860596,
+      "learning_rate": 4.660851176315094e-05,
+      "loss": 1.3245,
+      "step": 12840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0546751022338867,
+      "learning_rate": 4.660824742268041e-05,
+      "loss": 1.4496,
+      "step": 12841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7412937879562378,
+      "learning_rate": 4.6607983082209886e-05,
+      "loss": 1.489,
+      "step": 12842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9170500040054321,
+      "learning_rate": 4.6607718741739366e-05,
+      "loss": 1.4249,
+      "step": 12843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5371203422546387,
+      "learning_rate": 4.660745440126883e-05,
+      "loss": 1.3397,
+      "step": 12844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7472281455993652,
+      "learning_rate": 4.660719006079831e-05,
+      "loss": 1.4602,
+      "step": 12845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5353033542633057,
+      "learning_rate": 4.660692572032778e-05,
+      "loss": 1.2931,
+      "step": 12846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1383919715881348,
+      "learning_rate": 4.660666137985726e-05,
+      "loss": 1.307,
+      "step": 12847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6183024644851685,
+      "learning_rate": 4.6606397039386734e-05,
+      "loss": 1.3365,
+      "step": 12848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6902997493743896,
+      "learning_rate": 4.660613269891621e-05,
+      "loss": 1.403,
+      "step": 12849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6839790344238281,
+      "learning_rate": 4.660586835844568e-05,
+      "loss": 1.2995,
+      "step": 12850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6891446113586426,
+      "learning_rate": 4.6605604017975155e-05,
+      "loss": 1.4708,
+      "step": 12851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7937253713607788,
+      "learning_rate": 4.660533967750463e-05,
+      "loss": 1.5382,
+      "step": 12852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7405039072036743,
+      "learning_rate": 4.66050753370341e-05,
+      "loss": 1.6057,
+      "step": 12853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2064018249511719,
+      "learning_rate": 4.6604810996563575e-05,
+      "loss": 1.737,
+      "step": 12854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3894668817520142,
+      "learning_rate": 4.660454665609305e-05,
+      "loss": 1.1864,
+      "step": 12855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5374267101287842,
+      "learning_rate": 4.660428231562252e-05,
+      "loss": 1.3661,
+      "step": 12856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7393078804016113,
+      "learning_rate": 4.6604017975151996e-05,
+      "loss": 2.4931,
+      "step": 12857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4775371551513672,
+      "learning_rate": 4.6603753634681476e-05,
+      "loss": 1.3687,
+      "step": 12858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6358516216278076,
+      "learning_rate": 4.660348929421094e-05,
+      "loss": 1.6924,
+      "step": 12859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7748430967330933,
+      "learning_rate": 4.6603224953740424e-05,
+      "loss": 1.3124,
+      "step": 12860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3868916034698486,
+      "learning_rate": 4.660296061326989e-05,
+      "loss": 1.4973,
+      "step": 12861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5209600925445557,
+      "learning_rate": 4.660269627279937e-05,
+      "loss": 1.3656,
+      "step": 12862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6245801448822021,
+      "learning_rate": 4.6602431932328844e-05,
+      "loss": 1.3305,
+      "step": 12863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8289289474487305,
+      "learning_rate": 4.660216759185832e-05,
+      "loss": 1.2864,
+      "step": 12864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.574723482131958,
+      "learning_rate": 4.660190325138779e-05,
+      "loss": 1.3809,
+      "step": 12865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4059809446334839,
+      "learning_rate": 4.660163891091726e-05,
+      "loss": 1.291,
+      "step": 12866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5546528100967407,
+      "learning_rate": 4.660137457044674e-05,
+      "loss": 1.3339,
+      "step": 12867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.416063666343689,
+      "learning_rate": 4.660111022997621e-05,
+      "loss": 1.3209,
+      "step": 12868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0836427211761475,
+      "learning_rate": 4.6600845889505686e-05,
+      "loss": 1.428,
+      "step": 12869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7162907123565674,
+      "learning_rate": 4.660058154903516e-05,
+      "loss": 1.3029,
+      "step": 12870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7817026376724243,
+      "learning_rate": 4.660031720856463e-05,
+      "loss": 1.4421,
+      "step": 12871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7928714752197266,
+      "learning_rate": 4.6600052868094106e-05,
+      "loss": 1.4334,
+      "step": 12872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1574766635894775,
+      "learning_rate": 4.659978852762359e-05,
+      "loss": 1.3745,
+      "step": 12873
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7259286642074585,
+      "learning_rate": 4.6599524187153053e-05,
+      "loss": 1.3944,
+      "step": 12874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7041441202163696,
+      "learning_rate": 4.659925984668253e-05,
+      "loss": 1.3082,
+      "step": 12875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7457548379898071,
+      "learning_rate": 4.6598995506212e-05,
+      "loss": 1.4771,
+      "step": 12876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2478671073913574,
+      "learning_rate": 4.6598731165741474e-05,
+      "loss": 1.4417,
+      "step": 12877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9443464279174805,
+      "learning_rate": 4.6598466825270955e-05,
+      "loss": 1.4628,
+      "step": 12878
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9815884828567505,
+      "learning_rate": 4.659820248480042e-05,
+      "loss": 1.2766,
+      "step": 12879
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6087844371795654,
+      "learning_rate": 4.65979381443299e-05,
+      "loss": 1.2358,
+      "step": 12880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9717259407043457,
+      "learning_rate": 4.659767380385937e-05,
+      "loss": 1.4091,
+      "step": 12881
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.98126220703125,
+      "learning_rate": 4.659740946338885e-05,
+      "loss": 1.4576,
+      "step": 12882
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.633670687675476,
+      "learning_rate": 4.659714512291832e-05,
+      "loss": 2.0776,
+      "step": 12883
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9626044034957886,
+      "learning_rate": 4.6596880782447796e-05,
+      "loss": 1.2933,
+      "step": 12884
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2576279640197754,
+      "learning_rate": 4.659661644197727e-05,
+      "loss": 1.3476,
+      "step": 12885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7704358100891113,
+      "learning_rate": 4.659635210150674e-05,
+      "loss": 1.3638,
+      "step": 12886
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7906291484832764,
+      "learning_rate": 4.659608776103622e-05,
+      "loss": 1.5106,
+      "step": 12887
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.346308708190918,
+      "learning_rate": 4.659582342056569e-05,
+      "loss": 1.3488,
+      "step": 12888
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3551864624023438,
+      "learning_rate": 4.6595559080095164e-05,
+      "loss": 1.4678,
+      "step": 12889
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.933205485343933,
+      "learning_rate": 4.659529473962464e-05,
+      "loss": 1.2143,
+      "step": 12890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7415571212768555,
+      "learning_rate": 4.659503039915411e-05,
+      "loss": 1.5581,
+      "step": 12891
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.439751625061035,
+      "learning_rate": 4.6594766058683585e-05,
+      "loss": 1.5245,
+      "step": 12892
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9678690433502197,
+      "learning_rate": 4.6594501718213065e-05,
+      "loss": 1.5091,
+      "step": 12893
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9763144254684448,
+      "learning_rate": 4.659423737774253e-05,
+      "loss": 1.5285,
+      "step": 12894
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.615738868713379,
+      "learning_rate": 4.659397303727201e-05,
+      "loss": 1.3947,
+      "step": 12895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5145529508590698,
+      "learning_rate": 4.659370869680148e-05,
+      "loss": 1.3195,
+      "step": 12896
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8968948125839233,
+      "learning_rate": 4.659344435633096e-05,
+      "loss": 1.4499,
+      "step": 12897
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5287834405899048,
+      "learning_rate": 4.659318001586043e-05,
+      "loss": 1.3429,
+      "step": 12898
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4793319702148438,
+      "learning_rate": 4.6592915675389906e-05,
+      "loss": 1.272,
+      "step": 12899
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6604081392288208,
+      "learning_rate": 4.659265133491938e-05,
+      "loss": 1.3594,
+      "step": 12900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5374948978424072,
+      "learning_rate": 4.659238699444885e-05,
+      "loss": 1.3919,
+      "step": 12901
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9210193157196045,
+      "learning_rate": 4.659212265397833e-05,
+      "loss": 1.3705,
+      "step": 12902
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8143246173858643,
+      "learning_rate": 4.65918583135078e-05,
+      "loss": 1.3622,
+      "step": 12903
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3711652755737305,
+      "learning_rate": 4.6591593973037274e-05,
+      "loss": 1.3645,
+      "step": 12904
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.197852611541748,
+      "learning_rate": 4.659132963256675e-05,
+      "loss": 1.5032,
+      "step": 12905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8226454257965088,
+      "learning_rate": 4.659106529209622e-05,
+      "loss": 1.3399,
+      "step": 12906
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9004510641098022,
+      "learning_rate": 4.6590800951625695e-05,
+      "loss": 1.3455,
+      "step": 12907
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5385161638259888,
+      "learning_rate": 4.6590536611155175e-05,
+      "loss": 1.3065,
+      "step": 12908
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.103571653366089,
+      "learning_rate": 4.659027227068464e-05,
+      "loss": 1.4387,
+      "step": 12909
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.152658224105835,
+      "learning_rate": 4.659000793021412e-05,
+      "loss": 1.4601,
+      "step": 12910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8292406797409058,
+      "learning_rate": 4.658974358974359e-05,
+      "loss": 1.4052,
+      "step": 12911
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3784396648406982,
+      "learning_rate": 4.658947924927306e-05,
+      "loss": 1.1773,
+      "step": 12912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7966325283050537,
+      "learning_rate": 4.658921490880254e-05,
+      "loss": 1.3833,
+      "step": 12913
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8614429235458374,
+      "learning_rate": 4.658895056833201e-05,
+      "loss": 1.2761,
+      "step": 12914
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9443246126174927,
+      "learning_rate": 4.658868622786149e-05,
+      "loss": 1.5422,
+      "step": 12915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7233784198760986,
+      "learning_rate": 4.658842188739096e-05,
+      "loss": 1.4187,
+      "step": 12916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6504608392715454,
+      "learning_rate": 4.658815754692044e-05,
+      "loss": 1.2147,
+      "step": 12917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6291275024414062,
+      "learning_rate": 4.658789320644991e-05,
+      "loss": 1.1736,
+      "step": 12918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7254120111465454,
+      "learning_rate": 4.6587628865979385e-05,
+      "loss": 1.5187,
+      "step": 12919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6670023202896118,
+      "learning_rate": 4.658736452550886e-05,
+      "loss": 1.2924,
+      "step": 12920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5769898891448975,
+      "learning_rate": 4.658710018503833e-05,
+      "loss": 1.6109,
+      "step": 12921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7425624132156372,
+      "learning_rate": 4.6586835844567805e-05,
+      "loss": 1.413,
+      "step": 12922
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.722225546836853,
+      "learning_rate": 4.658657150409728e-05,
+      "loss": 1.5163,
+      "step": 12923
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5596681833267212,
+      "learning_rate": 4.658630716362675e-05,
+      "loss": 1.4598,
+      "step": 12924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.547309160232544,
+      "learning_rate": 4.6586042823156226e-05,
+      "loss": 1.3513,
+      "step": 12925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1612064838409424,
+      "learning_rate": 4.65857784826857e-05,
+      "loss": 1.2951,
+      "step": 12926
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5140504837036133,
+      "learning_rate": 4.658551414221517e-05,
+      "loss": 1.2313,
+      "step": 12927
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.874329924583435,
+      "learning_rate": 4.6585249801744653e-05,
+      "loss": 1.4642,
+      "step": 12928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8182315826416016,
+      "learning_rate": 4.658498546127412e-05,
+      "loss": 1.5511,
+      "step": 12929
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.178044557571411,
+      "learning_rate": 4.65847211208036e-05,
+      "loss": 1.5687,
+      "step": 12930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5581398010253906,
+      "learning_rate": 4.658445678033307e-05,
+      "loss": 1.4001,
+      "step": 12931
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4787882566452026,
+      "learning_rate": 4.658419243986255e-05,
+      "loss": 1.3734,
+      "step": 12932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5659650564193726,
+      "learning_rate": 4.658392809939202e-05,
+      "loss": 2.1921,
+      "step": 12933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1832475662231445,
+      "learning_rate": 4.6583663758921495e-05,
+      "loss": 1.5634,
+      "step": 12934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5424593687057495,
+      "learning_rate": 4.658339941845097e-05,
+      "loss": 1.3055,
+      "step": 12935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8207429647445679,
+      "learning_rate": 4.658313507798044e-05,
+      "loss": 1.3249,
+      "step": 12936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.031161069869995,
+      "learning_rate": 4.6582870737509916e-05,
+      "loss": 1.4501,
+      "step": 12937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1485836505889893,
+      "learning_rate": 4.658260639703939e-05,
+      "loss": 1.3801,
+      "step": 12938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0962183475494385,
+      "learning_rate": 4.658234205656886e-05,
+      "loss": 1.3889,
+      "step": 12939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8170398473739624,
+      "learning_rate": 4.6582077716098336e-05,
+      "loss": 1.4463,
+      "step": 12940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.933608055114746,
+      "learning_rate": 4.658181337562781e-05,
+      "loss": 1.3938,
+      "step": 12941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0020523071289062,
+      "learning_rate": 4.6581549035157283e-05,
+      "loss": 1.4714,
+      "step": 12942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8302679061889648,
+      "learning_rate": 4.6581284694686764e-05,
+      "loss": 1.4869,
+      "step": 12943
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3109450340270996,
+      "learning_rate": 4.658102035421623e-05,
+      "loss": 1.4851,
+      "step": 12944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6741610765457153,
+      "learning_rate": 4.658075601374571e-05,
+      "loss": 1.43,
+      "step": 12945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.089268445968628,
+      "learning_rate": 4.658049167327518e-05,
+      "loss": 1.5304,
+      "step": 12946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4824687242507935,
+      "learning_rate": 4.658022733280465e-05,
+      "loss": 1.2592,
+      "step": 12947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.108459234237671,
+      "learning_rate": 4.657996299233413e-05,
+      "loss": 1.5044,
+      "step": 12948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7880902290344238,
+      "learning_rate": 4.65796986518636e-05,
+      "loss": 1.472,
+      "step": 12949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8475145101547241,
+      "learning_rate": 4.657943431139308e-05,
+      "loss": 1.4053,
+      "step": 12950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.241864562034607,
+      "learning_rate": 4.6579169970922546e-05,
+      "loss": 1.6086,
+      "step": 12951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7136645317077637,
+      "learning_rate": 4.6578905630452026e-05,
+      "loss": 1.3567,
+      "step": 12952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4381489753723145,
+      "learning_rate": 4.65786412899815e-05,
+      "loss": 1.3095,
+      "step": 12953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4481501579284668,
+      "learning_rate": 4.657837694951097e-05,
+      "loss": 1.2879,
+      "step": 12954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1761488914489746,
+      "learning_rate": 4.657811260904045e-05,
+      "loss": 1.5326,
+      "step": 12955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6283408403396606,
+      "learning_rate": 4.657784826856992e-05,
+      "loss": 1.2737,
+      "step": 12956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5861947536468506,
+      "learning_rate": 4.6577583928099394e-05,
+      "loss": 1.3678,
+      "step": 12957
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.382094383239746,
+      "learning_rate": 4.657731958762887e-05,
+      "loss": 1.2441,
+      "step": 12958
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.803386926651001,
+      "learning_rate": 4.657705524715834e-05,
+      "loss": 1.5274,
+      "step": 12959
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9373620748519897,
+      "learning_rate": 4.6576790906687814e-05,
+      "loss": 1.4224,
+      "step": 12960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6900224685668945,
+      "learning_rate": 4.657652656621729e-05,
+      "loss": 1.2984,
+      "step": 12961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.345728874206543,
+      "learning_rate": 4.657626222574676e-05,
+      "loss": 1.3492,
+      "step": 12962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6804810762405396,
+      "learning_rate": 4.657599788527624e-05,
+      "loss": 1.4191,
+      "step": 12963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7946528196334839,
+      "learning_rate": 4.657573354480571e-05,
+      "loss": 1.5073,
+      "step": 12964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6645445823669434,
+      "learning_rate": 4.657546920433519e-05,
+      "loss": 1.2538,
+      "step": 12965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6943907737731934,
+      "learning_rate": 4.6575204863864656e-05,
+      "loss": 1.332,
+      "step": 12966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5033962726593018,
+      "learning_rate": 4.6574940523394136e-05,
+      "loss": 1.3165,
+      "step": 12967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109657049179077,
+      "learning_rate": 4.657467618292361e-05,
+      "loss": 1.3902,
+      "step": 12968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4155082702636719,
+      "learning_rate": 4.6574411842453083e-05,
+      "loss": 1.2479,
+      "step": 12969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9534001350402832,
+      "learning_rate": 4.657414750198256e-05,
+      "loss": 1.6105,
+      "step": 12970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4861834049224854,
+      "learning_rate": 4.657388316151203e-05,
+      "loss": 1.5812,
+      "step": 12971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.699104905128479,
+      "learning_rate": 4.6573618821041504e-05,
+      "loss": 1.2994,
+      "step": 12972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1496458053588867,
+      "learning_rate": 4.657335448057098e-05,
+      "loss": 1.3756,
+      "step": 12973
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5350924730300903,
+      "learning_rate": 4.657309014010045e-05,
+      "loss": 1.3482,
+      "step": 12974
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3078527450561523,
+      "learning_rate": 4.6572825799629925e-05,
+      "loss": 1.5083,
+      "step": 12975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8589510917663574,
+      "learning_rate": 4.65725614591594e-05,
+      "loss": 1.3139,
+      "step": 12976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6810287237167358,
+      "learning_rate": 4.657229711868887e-05,
+      "loss": 1.3123,
+      "step": 12977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.819278359413147,
+      "learning_rate": 4.657203277821835e-05,
+      "loss": 1.5006,
+      "step": 12978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7298088073730469,
+      "learning_rate": 4.657176843774782e-05,
+      "loss": 1.2258,
+      "step": 12979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5903468132019043,
+      "learning_rate": 4.65715040972773e-05,
+      "loss": 1.5086,
+      "step": 12980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9251362085342407,
+      "learning_rate": 4.6571239756806766e-05,
+      "loss": 1.3595,
+      "step": 12981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.124802350997925,
+      "learning_rate": 4.6570975416336247e-05,
+      "loss": 1.4962,
+      "step": 12982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.717292308807373,
+      "learning_rate": 4.657071107586572e-05,
+      "loss": 1.4369,
+      "step": 12983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.580222249031067,
+      "learning_rate": 4.657044673539519e-05,
+      "loss": 2.1058,
+      "step": 12984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7885661125183105,
+      "learning_rate": 4.657018239492467e-05,
+      "loss": 1.4244,
+      "step": 12985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9960911273956299,
+      "learning_rate": 4.6569918054454134e-05,
+      "loss": 1.4372,
+      "step": 12986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5257043838500977,
+      "learning_rate": 4.6569653713983614e-05,
+      "loss": 1.3622,
+      "step": 12987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.059953212738037,
+      "learning_rate": 4.656938937351309e-05,
+      "loss": 1.4405,
+      "step": 12988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.048341989517212,
+      "learning_rate": 4.656912503304256e-05,
+      "loss": 1.3966,
+      "step": 12989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5396535396575928,
+      "learning_rate": 4.6568860692572035e-05,
+      "loss": 1.5421,
+      "step": 12990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5633467435836792,
+      "learning_rate": 4.656859635210151e-05,
+      "loss": 1.3311,
+      "step": 12991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.699035882949829,
+      "learning_rate": 4.656833201163098e-05,
+      "loss": 1.2112,
+      "step": 12992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8690966367721558,
+      "learning_rate": 4.6568067671160456e-05,
+      "loss": 1.324,
+      "step": 12993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.075730085372925,
+      "learning_rate": 4.656780333068993e-05,
+      "loss": 1.4394,
+      "step": 12994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5270289182662964,
+      "learning_rate": 4.65675389902194e-05,
+      "loss": 1.3581,
+      "step": 12995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4154754877090454,
+      "learning_rate": 4.6567274649748877e-05,
+      "loss": 1.2432,
+      "step": 12996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6977665424346924,
+      "learning_rate": 4.656701030927835e-05,
+      "loss": 1.2968,
+      "step": 12997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.319898009300232,
+      "learning_rate": 4.656674596880783e-05,
+      "loss": 1.0739,
+      "step": 12998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8038805723190308,
+      "learning_rate": 4.65664816283373e-05,
+      "loss": 1.3145,
+      "step": 12999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6329327821731567,
+      "learning_rate": 4.656621728786678e-05,
+      "loss": 1.3551,
+      "step": 13000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7554391622543335,
+      "learning_rate": 4.6565952947396244e-05,
+      "loss": 1.254,
+      "step": 13001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5699172019958496,
+      "learning_rate": 4.6565688606925725e-05,
+      "loss": 1.3284,
+      "step": 13002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9797130823135376,
+      "learning_rate": 4.65654242664552e-05,
+      "loss": 1.406,
+      "step": 13003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5314728021621704,
+      "learning_rate": 4.656515992598467e-05,
+      "loss": 1.1442,
+      "step": 13004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7122243642807007,
+      "learning_rate": 4.6564895585514145e-05,
+      "loss": 1.3824,
+      "step": 13005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.261949062347412,
+      "learning_rate": 4.656463124504362e-05,
+      "loss": 1.3192,
+      "step": 13006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8349401950836182,
+      "learning_rate": 4.656436690457309e-05,
+      "loss": 1.2554,
+      "step": 13007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6803287267684937,
+      "learning_rate": 4.6564102564102566e-05,
+      "loss": 1.2669,
+      "step": 13008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7419521808624268,
+      "learning_rate": 4.656383822363204e-05,
+      "loss": 1.3202,
+      "step": 13009
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1769988536834717,
+      "learning_rate": 4.656357388316151e-05,
+      "loss": 1.4643,
+      "step": 13010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9064397811889648,
+      "learning_rate": 4.656330954269099e-05,
+      "loss": 1.259,
+      "step": 13011
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8620882034301758,
+      "learning_rate": 4.656304520222046e-05,
+      "loss": 1.2225,
+      "step": 13012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0476016998291016,
+      "learning_rate": 4.656278086174994e-05,
+      "loss": 1.4767,
+      "step": 13013
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0405476093292236,
+      "learning_rate": 4.656251652127941e-05,
+      "loss": 1.2664,
+      "step": 13014
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1758573055267334,
+      "learning_rate": 4.656225218080889e-05,
+      "loss": 1.3994,
+      "step": 13015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.301055908203125,
+      "learning_rate": 4.6561987840338355e-05,
+      "loss": 1.4245,
+      "step": 13016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6869292259216309,
+      "learning_rate": 4.6561723499867835e-05,
+      "loss": 1.3733,
+      "step": 13017
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.414988398551941,
+      "learning_rate": 4.656145915939731e-05,
+      "loss": 1.1677,
+      "step": 13018
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8044565916061401,
+      "learning_rate": 4.6561194818926775e-05,
+      "loss": 1.2884,
+      "step": 13019
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1854429244995117,
+      "learning_rate": 4.6560930478456256e-05,
+      "loss": 1.4416,
+      "step": 13020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7631641626358032,
+      "learning_rate": 4.656066613798572e-05,
+      "loss": 1.3586,
+      "step": 13021
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3194830417633057,
+      "learning_rate": 4.65604017975152e-05,
+      "loss": 1.4016,
+      "step": 13022
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.970513105392456,
+      "learning_rate": 4.6560137457044677e-05,
+      "loss": 1.3493,
+      "step": 13023
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5418251752853394,
+      "learning_rate": 4.655987311657415e-05,
+      "loss": 1.2854,
+      "step": 13024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.416574001312256,
+      "learning_rate": 4.6559608776103624e-05,
+      "loss": 1.5659,
+      "step": 13025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.642543077468872,
+      "learning_rate": 4.65593444356331e-05,
+      "loss": 1.2868,
+      "step": 13026
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.22430157661438,
+      "learning_rate": 4.655908009516257e-05,
+      "loss": 1.4426,
+      "step": 13027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8639800548553467,
+      "learning_rate": 4.655881575469205e-05,
+      "loss": 1.3128,
+      "step": 13028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4395463466644287,
+      "learning_rate": 4.655855141422152e-05,
+      "loss": 1.4696,
+      "step": 13029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7927825450897217,
+      "learning_rate": 4.655828707375099e-05,
+      "loss": 1.5151,
+      "step": 13030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.561674952507019,
+      "learning_rate": 4.6558022733280465e-05,
+      "loss": 2.3129,
+      "step": 13031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.071855306625366,
+      "learning_rate": 4.655775839280994e-05,
+      "loss": 1.4796,
+      "step": 13032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5406659841537476,
+      "learning_rate": 4.655749405233942e-05,
+      "loss": 1.4701,
+      "step": 13033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8174089193344116,
+      "learning_rate": 4.6557229711868886e-05,
+      "loss": 1.4205,
+      "step": 13034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.143979787826538,
+      "learning_rate": 4.6556965371398366e-05,
+      "loss": 1.3989,
+      "step": 13035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7232080698013306,
+      "learning_rate": 4.655670103092783e-05,
+      "loss": 1.2787,
+      "step": 13036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.371715545654297,
+      "learning_rate": 4.655643669045731e-05,
+      "loss": 1.499,
+      "step": 13037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5378042459487915,
+      "learning_rate": 4.655617234998679e-05,
+      "loss": 1.3278,
+      "step": 13038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.071113348007202,
+      "learning_rate": 4.655590800951626e-05,
+      "loss": 1.5207,
+      "step": 13039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.933044672012329,
+      "learning_rate": 4.6555643669045734e-05,
+      "loss": 1.3559,
+      "step": 13040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8555974960327148,
+      "learning_rate": 4.655537932857521e-05,
+      "loss": 1.484,
+      "step": 13041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.660544991493225,
+      "learning_rate": 4.655511498810468e-05,
+      "loss": 1.2966,
+      "step": 13042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5812093019485474,
+      "learning_rate": 4.6554850647634155e-05,
+      "loss": 1.2011,
+      "step": 13043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.870549201965332,
+      "learning_rate": 4.655458630716363e-05,
+      "loss": 1.3199,
+      "step": 13044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6115550994873047,
+      "learning_rate": 4.65543219666931e-05,
+      "loss": 1.2536,
+      "step": 13045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4938488006591797,
+      "learning_rate": 4.6554057626222575e-05,
+      "loss": 1.2234,
+      "step": 13046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9533089399337769,
+      "learning_rate": 4.655379328575205e-05,
+      "loss": 1.4155,
+      "step": 13047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.554121494293213,
+      "learning_rate": 4.655352894528153e-05,
+      "loss": 1.4988,
+      "step": 13048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.61648428440094,
+      "learning_rate": 4.6553264604810996e-05,
+      "loss": 1.2988,
+      "step": 13049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8977054357528687,
+      "learning_rate": 4.6553000264340477e-05,
+      "loss": 1.4896,
+      "step": 13050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6491881608963013,
+      "learning_rate": 4.655273592386994e-05,
+      "loss": 1.24,
+      "step": 13051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8698722124099731,
+      "learning_rate": 4.6552471583399424e-05,
+      "loss": 1.3829,
+      "step": 13052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4878215789794922,
+      "learning_rate": 4.65522072429289e-05,
+      "loss": 1.3901,
+      "step": 13053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5383615493774414,
+      "learning_rate": 4.655194290245837e-05,
+      "loss": 1.3346,
+      "step": 13054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7889909744262695,
+      "learning_rate": 4.6551678561987844e-05,
+      "loss": 1.2791,
+      "step": 13055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5573855638504028,
+      "learning_rate": 4.655141422151731e-05,
+      "loss": 1.255,
+      "step": 13056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0462892055511475,
+      "learning_rate": 4.655114988104679e-05,
+      "loss": 1.4823,
+      "step": 13057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9634896516799927,
+      "learning_rate": 4.6550885540576265e-05,
+      "loss": 1.4084,
+      "step": 13058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.642160177230835,
+      "learning_rate": 4.655062120010574e-05,
+      "loss": 1.2931,
+      "step": 13059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5965549945831299,
+      "learning_rate": 4.655035685963521e-05,
+      "loss": 1.367,
+      "step": 13060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.334486484527588,
+      "learning_rate": 4.6550092519164686e-05,
+      "loss": 1.548,
+      "step": 13061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7457321882247925,
+      "learning_rate": 4.654982817869416e-05,
+      "loss": 1.4204,
+      "step": 13062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7607413530349731,
+      "learning_rate": 4.654956383822364e-05,
+      "loss": 1.4958,
+      "step": 13063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1367428302764893,
+      "learning_rate": 4.6549299497753107e-05,
+      "loss": 1.5656,
+      "step": 13064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8022069931030273,
+      "learning_rate": 4.654903515728258e-05,
+      "loss": 1.5261,
+      "step": 13065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6428107023239136,
+      "learning_rate": 4.6548770816812054e-05,
+      "loss": 1.4169,
+      "step": 13066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.551151156425476,
+      "learning_rate": 4.654850647634153e-05,
+      "loss": 1.3246,
+      "step": 13067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2521464824676514,
+      "learning_rate": 4.654824213587101e-05,
+      "loss": 1.3737,
+      "step": 13068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0327723026275635,
+      "learning_rate": 4.6547977795400474e-05,
+      "loss": 1.5088,
+      "step": 13069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7939807176589966,
+      "learning_rate": 4.6547713454929955e-05,
+      "loss": 1.3441,
+      "step": 13070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8435614109039307,
+      "learning_rate": 4.654744911445942e-05,
+      "loss": 2.0825,
+      "step": 13071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5436453819274902,
+      "learning_rate": 4.65471847739889e-05,
+      "loss": 1.6751,
+      "step": 13072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0471994876861572,
+      "learning_rate": 4.6546920433518375e-05,
+      "loss": 1.3976,
+      "step": 13073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.247422933578491,
+      "learning_rate": 4.654665609304785e-05,
+      "loss": 1.4598,
+      "step": 13074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9363102912902832,
+      "learning_rate": 4.654639175257732e-05,
+      "loss": 1.401,
+      "step": 13075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.077181100845337,
+      "learning_rate": 4.6546127412106796e-05,
+      "loss": 1.5628,
+      "step": 13076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8894965648651123,
+      "learning_rate": 4.654586307163627e-05,
+      "loss": 2.1241,
+      "step": 13077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5036571025848389,
+      "learning_rate": 4.654559873116574e-05,
+      "loss": 1.1743,
+      "step": 13078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1500866413116455,
+      "learning_rate": 4.654533439069522e-05,
+      "loss": 1.424,
+      "step": 13079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.251657724380493,
+      "learning_rate": 4.654507005022469e-05,
+      "loss": 1.3392,
+      "step": 13080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9304335117340088,
+      "learning_rate": 4.6544805709754164e-05,
+      "loss": 1.5332,
+      "step": 13081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5704982280731201,
+      "learning_rate": 4.654454136928364e-05,
+      "loss": 2.2367,
+      "step": 13082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2256815433502197,
+      "learning_rate": 4.654427702881312e-05,
+      "loss": 1.5316,
+      "step": 13083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9833179712295532,
+      "learning_rate": 4.6544012688342585e-05,
+      "loss": 1.3112,
+      "step": 13084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0555167198181152,
+      "learning_rate": 4.6543748347872065e-05,
+      "loss": 2.6033,
+      "step": 13085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9138455390930176,
+      "learning_rate": 4.654348400740153e-05,
+      "loss": 1.3783,
+      "step": 13086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4011471271514893,
+      "learning_rate": 4.654321966693101e-05,
+      "loss": 1.3187,
+      "step": 13087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2588744163513184,
+      "learning_rate": 4.6542955326460486e-05,
+      "loss": 1.4146,
+      "step": 13088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.591955542564392,
+      "learning_rate": 4.654269098598996e-05,
+      "loss": 1.3915,
+      "step": 13089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8884432315826416,
+      "learning_rate": 4.654242664551943e-05,
+      "loss": 1.3999,
+      "step": 13090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8019579648971558,
+      "learning_rate": 4.65421623050489e-05,
+      "loss": 1.2465,
+      "step": 13091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8571714162826538,
+      "learning_rate": 4.654189796457838e-05,
+      "loss": 1.3781,
+      "step": 13092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.139694929122925,
+      "learning_rate": 4.6541633624107854e-05,
+      "loss": 1.4509,
+      "step": 13093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5766901969909668,
+      "learning_rate": 4.654136928363733e-05,
+      "loss": 1.3244,
+      "step": 13094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9077904224395752,
+      "learning_rate": 4.65411049431668e-05,
+      "loss": 1.5086,
+      "step": 13095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5451289415359497,
+      "learning_rate": 4.6540840602696274e-05,
+      "loss": 1.3438,
+      "step": 13096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6879903078079224,
+      "learning_rate": 4.654057626222575e-05,
+      "loss": 1.4047,
+      "step": 13097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9306191205978394,
+      "learning_rate": 4.654031192175523e-05,
+      "loss": 1.354,
+      "step": 13098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9255198240280151,
+      "learning_rate": 4.6540047581284695e-05,
+      "loss": 1.4079,
+      "step": 13099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1509714126586914,
+      "learning_rate": 4.6539783240814175e-05,
+      "loss": 1.4823,
+      "step": 13100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0126702785491943,
+      "learning_rate": 4.653951890034364e-05,
+      "loss": 2.1413,
+      "step": 13101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3850739002227783,
+      "learning_rate": 4.6539254559873116e-05,
+      "loss": 1.4908,
+      "step": 13102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.728318214416504,
+      "learning_rate": 4.6538990219402596e-05,
+      "loss": 1.4694,
+      "step": 13103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4787862300872803,
+      "learning_rate": 4.653872587893206e-05,
+      "loss": 1.5361,
+      "step": 13104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9491418600082397,
+      "learning_rate": 4.653846153846154e-05,
+      "loss": 1.359,
+      "step": 13105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1073944568634033,
+      "learning_rate": 4.653819719799101e-05,
+      "loss": 1.4287,
+      "step": 13106
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6963034868240356,
+      "learning_rate": 4.653793285752049e-05,
+      "loss": 1.4304,
+      "step": 13107
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.146219253540039,
+      "learning_rate": 4.6537668517049964e-05,
+      "loss": 1.3977,
+      "step": 13108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5413070917129517,
+      "learning_rate": 4.653740417657944e-05,
+      "loss": 1.3195,
+      "step": 13109
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7755670547485352,
+      "learning_rate": 4.653713983610891e-05,
+      "loss": 1.3561,
+      "step": 13110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8577306270599365,
+      "learning_rate": 4.6536875495638385e-05,
+      "loss": 1.3964,
+      "step": 13111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9136618375778198,
+      "learning_rate": 4.653661115516786e-05,
+      "loss": 1.3111,
+      "step": 13112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.774796724319458,
+      "learning_rate": 4.653634681469733e-05,
+      "loss": 1.421,
+      "step": 13113
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7632765769958496,
+      "learning_rate": 4.6536082474226805e-05,
+      "loss": 1.389,
+      "step": 13114
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4369056224823,
+      "learning_rate": 4.653581813375628e-05,
+      "loss": 1.4563,
+      "step": 13115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2030601501464844,
+      "learning_rate": 4.653555379328575e-05,
+      "loss": 1.5029,
+      "step": 13116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7027677297592163,
+      "learning_rate": 4.6535289452815226e-05,
+      "loss": 1.353,
+      "step": 13117
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9637149572372437,
+      "learning_rate": 4.6535025112344706e-05,
+      "loss": 1.5361,
+      "step": 13118
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6581487655639648,
+      "learning_rate": 4.653476077187417e-05,
+      "loss": 1.2937,
+      "step": 13119
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7974815368652344,
+      "learning_rate": 4.6534496431403654e-05,
+      "loss": 1.4339,
+      "step": 13120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5358061790466309,
+      "learning_rate": 4.653423209093312e-05,
+      "loss": 1.3018,
+      "step": 13121
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.366190195083618,
+      "learning_rate": 4.65339677504626e-05,
+      "loss": 1.7186,
+      "step": 13122
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.351463794708252,
+      "learning_rate": 4.6533703409992074e-05,
+      "loss": 1.3544,
+      "step": 13123
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7752048969268799,
+      "learning_rate": 4.653343906952155e-05,
+      "loss": 1.1938,
+      "step": 13124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.614018201828003,
+      "learning_rate": 4.653317472905102e-05,
+      "loss": 1.3345,
+      "step": 13125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.636189579963684,
+      "learning_rate": 4.6532910388580495e-05,
+      "loss": 1.396,
+      "step": 13126
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7134909629821777,
+      "learning_rate": 4.653264604810997e-05,
+      "loss": 1.5311,
+      "step": 13127
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8548691272735596,
+      "learning_rate": 4.653238170763944e-05,
+      "loss": 1.4104,
+      "step": 13128
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.670939564704895,
+      "learning_rate": 4.6532117367168916e-05,
+      "loss": 1.3993,
+      "step": 13129
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.066041946411133,
+      "learning_rate": 4.653185302669839e-05,
+      "loss": 1.3899,
+      "step": 13130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8951152563095093,
+      "learning_rate": 4.653158868622786e-05,
+      "loss": 1.5088,
+      "step": 13131
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7337788343429565,
+      "learning_rate": 4.6531324345757336e-05,
+      "loss": 1.8799,
+      "step": 13132
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9485268592834473,
+      "learning_rate": 4.653106000528682e-05,
+      "loss": 1.3824,
+      "step": 13133
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6811141967773438,
+      "learning_rate": 4.6530795664816284e-05,
+      "loss": 1.4576,
+      "step": 13134
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3831114768981934,
+      "learning_rate": 4.6530531324345764e-05,
+      "loss": 1.3136,
+      "step": 13135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6714273691177368,
+      "learning_rate": 4.653026698387523e-05,
+      "loss": 1.4902,
+      "step": 13136
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.837250828742981,
+      "learning_rate": 4.6530002643404704e-05,
+      "loss": 1.3161,
+      "step": 13137
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.762759804725647,
+      "learning_rate": 4.6529738302934185e-05,
+      "loss": 1.3947,
+      "step": 13138
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.646599531173706,
+      "learning_rate": 4.652947396246365e-05,
+      "loss": 1.5616,
+      "step": 13139
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.110502243041992,
+      "learning_rate": 4.652920962199313e-05,
+      "loss": 1.3975,
+      "step": 13140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9684492349624634,
+      "learning_rate": 4.65289452815226e-05,
+      "loss": 1.5689,
+      "step": 13141
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4459645748138428,
+      "learning_rate": 4.652868094105208e-05,
+      "loss": 1.5707,
+      "step": 13142
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6337029933929443,
+      "learning_rate": 4.652841660058155e-05,
+      "loss": 1.3177,
+      "step": 13143
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6511789560317993,
+      "learning_rate": 4.6528152260111026e-05,
+      "loss": 1.3273,
+      "step": 13144
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6181145906448364,
+      "learning_rate": 4.65278879196405e-05,
+      "loss": 1.2453,
+      "step": 13145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.519029378890991,
+      "learning_rate": 4.652762357916997e-05,
+      "loss": 1.3903,
+      "step": 13146
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5838589668273926,
+      "learning_rate": 4.652735923869945e-05,
+      "loss": 1.2582,
+      "step": 13147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0296969413757324,
+      "learning_rate": 4.652709489822892e-05,
+      "loss": 2.6292,
+      "step": 13148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8268102407455444,
+      "learning_rate": 4.6526830557758394e-05,
+      "loss": 1.2643,
+      "step": 13149
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8726125955581665,
+      "learning_rate": 4.652656621728787e-05,
+      "loss": 1.4671,
+      "step": 13150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.371476888656616,
+      "learning_rate": 4.652630187681734e-05,
+      "loss": 1.4755,
+      "step": 13151
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4940292835235596,
+      "learning_rate": 4.6526037536346815e-05,
+      "loss": 1.3402,
+      "step": 13152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7729530334472656,
+      "learning_rate": 4.6525773195876295e-05,
+      "loss": 1.363,
+      "step": 13153
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7228758335113525,
+      "learning_rate": 4.652550885540576e-05,
+      "loss": 1.3471,
+      "step": 13154
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.606887936592102,
+      "learning_rate": 4.652524451493524e-05,
+      "loss": 1.464,
+      "step": 13155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0015554428100586,
+      "learning_rate": 4.652498017446471e-05,
+      "loss": 1.4947,
+      "step": 13156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6266220808029175,
+      "learning_rate": 4.652471583399419e-05,
+      "loss": 2.5123,
+      "step": 13157
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.983076810836792,
+      "learning_rate": 4.652445149352366e-05,
+      "loss": 1.3362,
+      "step": 13158
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8428889513015747,
+      "learning_rate": 4.6524187153053136e-05,
+      "loss": 1.4151,
+      "step": 13159
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9423627853393555,
+      "learning_rate": 4.652392281258261e-05,
+      "loss": 1.3872,
+      "step": 13160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9679332971572876,
+      "learning_rate": 4.6523658472112084e-05,
+      "loss": 1.4091,
+      "step": 13161
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7710978984832764,
+      "learning_rate": 4.652339413164156e-05,
+      "loss": 1.283,
+      "step": 13162
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.805640459060669,
+      "learning_rate": 4.652312979117103e-05,
+      "loss": 1.2091,
+      "step": 13163
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2952091693878174,
+      "learning_rate": 4.6522865450700504e-05,
+      "loss": 1.5728,
+      "step": 13164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8169612884521484,
+      "learning_rate": 4.652260111022998e-05,
+      "loss": 1.417,
+      "step": 13165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5399268865585327,
+      "learning_rate": 4.652233676975945e-05,
+      "loss": 1.4113,
+      "step": 13166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5784060955047607,
+      "learning_rate": 4.6522072429288925e-05,
+      "loss": 1.5166,
+      "step": 13167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.41334867477417,
+      "learning_rate": 4.6521808088818405e-05,
+      "loss": 1.2678,
+      "step": 13168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9667823314666748,
+      "learning_rate": 4.652154374834787e-05,
+      "loss": 1.3553,
+      "step": 13169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2802553176879883,
+      "learning_rate": 4.652127940787735e-05,
+      "loss": 1.4847,
+      "step": 13170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.911973237991333,
+      "learning_rate": 4.652101506740682e-05,
+      "loss": 1.4887,
+      "step": 13171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6293604373931885,
+      "learning_rate": 4.65207507269363e-05,
+      "loss": 1.3749,
+      "step": 13172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.889673948287964,
+      "learning_rate": 4.652048638646577e-05,
+      "loss": 1.5902,
+      "step": 13173
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8127223253250122,
+      "learning_rate": 4.652022204599524e-05,
+      "loss": 1.6018,
+      "step": 13174
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5838220119476318,
+      "learning_rate": 4.651995770552472e-05,
+      "loss": 1.9498,
+      "step": 13175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7459114789962769,
+      "learning_rate": 4.651969336505419e-05,
+      "loss": 1.3178,
+      "step": 13176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4429131746292114,
+      "learning_rate": 4.651942902458367e-05,
+      "loss": 1.2118,
+      "step": 13177
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6585917472839355,
+      "learning_rate": 4.651916468411314e-05,
+      "loss": 1.2843,
+      "step": 13178
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.806782841682434,
+      "learning_rate": 4.6518900343642615e-05,
+      "loss": 1.4403,
+      "step": 13179
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.004467487335205,
+      "learning_rate": 4.651863600317209e-05,
+      "loss": 1.4663,
+      "step": 13180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.905834436416626,
+      "learning_rate": 4.651837166270156e-05,
+      "loss": 1.3606,
+      "step": 13181
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7411525249481201,
+      "learning_rate": 4.6518107322231035e-05,
+      "loss": 1.4925,
+      "step": 13182
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.705895185470581,
+      "learning_rate": 4.651784298176051e-05,
+      "loss": 1.4639,
+      "step": 13183
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.27610182762146,
+      "learning_rate": 4.651757864128998e-05,
+      "loss": 1.4434,
+      "step": 13184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5673216581344604,
+      "learning_rate": 4.6517314300819456e-05,
+      "loss": 1.2347,
+      "step": 13185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.492250442504883,
+      "learning_rate": 4.651704996034893e-05,
+      "loss": 1.5947,
+      "step": 13186
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6387664079666138,
+      "learning_rate": 4.65167856198784e-05,
+      "loss": 1.3374,
+      "step": 13187
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5726059675216675,
+      "learning_rate": 4.6516521279407884e-05,
+      "loss": 1.4062,
+      "step": 13188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.917160749435425,
+      "learning_rate": 4.651625693893735e-05,
+      "loss": 1.4189,
+      "step": 13189
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.325792074203491,
+      "learning_rate": 4.651599259846683e-05,
+      "loss": 1.3891,
+      "step": 13190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.689364910125732,
+      "learning_rate": 4.65157282579963e-05,
+      "loss": 1.5563,
+      "step": 13191
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2184624671936035,
+      "learning_rate": 4.651546391752578e-05,
+      "loss": 1.5133,
+      "step": 13192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7763727903366089,
+      "learning_rate": 4.651519957705525e-05,
+      "loss": 1.2646,
+      "step": 13193
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.599881410598755,
+      "learning_rate": 4.6514935236584725e-05,
+      "loss": 1.4346,
+      "step": 13194
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8558803796768188,
+      "learning_rate": 4.65146708961142e-05,
+      "loss": 1.4977,
+      "step": 13195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8096979856491089,
+      "learning_rate": 4.651440655564367e-05,
+      "loss": 1.3652,
+      "step": 13196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.640392780303955,
+      "learning_rate": 4.6514142215173146e-05,
+      "loss": 1.3216,
+      "step": 13197
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8637559413909912,
+      "learning_rate": 4.651387787470262e-05,
+      "loss": 1.3887,
+      "step": 13198
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6489641666412354,
+      "learning_rate": 4.651361353423209e-05,
+      "loss": 1.3724,
+      "step": 13199
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5582607984542847,
+      "learning_rate": 4.6513349193761566e-05,
+      "loss": 1.3167,
+      "step": 13200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4617156982421875,
+      "learning_rate": 4.651308485329104e-05,
+      "loss": 1.2684,
+      "step": 13201
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8670783042907715,
+      "learning_rate": 4.6512820512820514e-05,
+      "loss": 1.4509,
+      "step": 13202
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.428677558898926,
+      "learning_rate": 4.6512556172349994e-05,
+      "loss": 1.5294,
+      "step": 13203
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6806020736694336,
+      "learning_rate": 4.651229183187946e-05,
+      "loss": 1.4079,
+      "step": 13204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.588516354560852,
+      "learning_rate": 4.651202749140894e-05,
+      "loss": 1.1903,
+      "step": 13205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9867103099822998,
+      "learning_rate": 4.651176315093841e-05,
+      "loss": 1.4737,
+      "step": 13206
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015059471130371,
+      "learning_rate": 4.651149881046789e-05,
+      "loss": 1.4372,
+      "step": 13207
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.301788568496704,
+      "learning_rate": 4.651123446999736e-05,
+      "loss": 1.5147,
+      "step": 13208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9610880613327026,
+      "learning_rate": 4.651097012952683e-05,
+      "loss": 1.3676,
+      "step": 13209
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1157076358795166,
+      "learning_rate": 4.651070578905631e-05,
+      "loss": 1.3861,
+      "step": 13210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5677810907363892,
+      "learning_rate": 4.6510441448585776e-05,
+      "loss": 1.1534,
+      "step": 13211
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8507052659988403,
+      "learning_rate": 4.6510177108115256e-05,
+      "loss": 1.4487,
+      "step": 13212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.351081132888794,
+      "learning_rate": 4.650991276764473e-05,
+      "loss": 1.2965,
+      "step": 13213
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6287940740585327,
+      "learning_rate": 4.65096484271742e-05,
+      "loss": 1.4021,
+      "step": 13214
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7770947217941284,
+      "learning_rate": 4.650938408670368e-05,
+      "loss": 1.9559,
+      "step": 13215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3883326053619385,
+      "learning_rate": 4.650911974623315e-05,
+      "loss": 1.2996,
+      "step": 13216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7529914379119873,
+      "learning_rate": 4.6508855405762624e-05,
+      "loss": 1.3198,
+      "step": 13217
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9912165403366089,
+      "learning_rate": 4.6508591065292104e-05,
+      "loss": 1.4341,
+      "step": 13218
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3362629413604736,
+      "learning_rate": 4.650832672482157e-05,
+      "loss": 1.3565,
+      "step": 13219
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.064723253250122,
+      "learning_rate": 4.6508062384351045e-05,
+      "loss": 1.2755,
+      "step": 13220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5744247436523438,
+      "learning_rate": 4.650779804388052e-05,
+      "loss": 1.2775,
+      "step": 13221
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.445124864578247,
+      "learning_rate": 4.650753370340999e-05,
+      "loss": 1.5063,
+      "step": 13222
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6202178001403809,
+      "learning_rate": 4.650726936293947e-05,
+      "loss": 1.3903,
+      "step": 13223
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5075005292892456,
+      "learning_rate": 4.650700502246894e-05,
+      "loss": 1.3219,
+      "step": 13224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03092885017395,
+      "learning_rate": 4.650674068199842e-05,
+      "loss": 1.3473,
+      "step": 13225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2109345197677612,
+      "learning_rate": 4.6506476341527886e-05,
+      "loss": 2.156,
+      "step": 13226
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.230736494064331,
+      "learning_rate": 4.6506212001057366e-05,
+      "loss": 1.6267,
+      "step": 13227
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9533820152282715,
+      "learning_rate": 4.650594766058684e-05,
+      "loss": 1.3946,
+      "step": 13228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.191777467727661,
+      "learning_rate": 4.6505683320116313e-05,
+      "loss": 1.4676,
+      "step": 13229
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.641740083694458,
+      "learning_rate": 4.650541897964579e-05,
+      "loss": 1.3436,
+      "step": 13230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.591820240020752,
+      "learning_rate": 4.650515463917526e-05,
+      "loss": 1.3415,
+      "step": 13231
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.678290605545044,
+      "learning_rate": 4.6504890298704734e-05,
+      "loss": 1.5102,
+      "step": 13232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8001490831375122,
+      "learning_rate": 4.650462595823421e-05,
+      "loss": 1.1537,
+      "step": 13233
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6598528623580933,
+      "learning_rate": 4.650436161776368e-05,
+      "loss": 1.5651,
+      "step": 13234
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.001244306564331,
+      "learning_rate": 4.6504097277293155e-05,
+      "loss": 1.5063,
+      "step": 13235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.468166470527649,
+      "learning_rate": 4.650383293682263e-05,
+      "loss": 1.2513,
+      "step": 13236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8545671701431274,
+      "learning_rate": 4.65035685963521e-05,
+      "loss": 1.421,
+      "step": 13237
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.215500593185425,
+      "learning_rate": 4.650330425588158e-05,
+      "loss": 1.5081,
+      "step": 13238
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3871889114379883,
+      "learning_rate": 4.650303991541105e-05,
+      "loss": 1.4709,
+      "step": 13239
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6007928848266602,
+      "learning_rate": 4.650277557494053e-05,
+      "loss": 1.4818,
+      "step": 13240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5169472694396973,
+      "learning_rate": 4.6502511234469996e-05,
+      "loss": 1.2798,
+      "step": 13241
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6629760265350342,
+      "learning_rate": 4.650224689399948e-05,
+      "loss": 1.302,
+      "step": 13242
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9559249877929688,
+      "learning_rate": 4.650198255352895e-05,
+      "loss": 1.386,
+      "step": 13243
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1335971355438232,
+      "learning_rate": 4.6501718213058424e-05,
+      "loss": 1.4528,
+      "step": 13244
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.529828667640686,
+      "learning_rate": 4.65014538725879e-05,
+      "loss": 1.3609,
+      "step": 13245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5953311920166016,
+      "learning_rate": 4.6501189532117364e-05,
+      "loss": 1.311,
+      "step": 13246
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171482563018799,
+      "learning_rate": 4.6500925191646845e-05,
+      "loss": 1.5457,
+      "step": 13247
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1190061569213867,
+      "learning_rate": 4.650066085117632e-05,
+      "loss": 1.5252,
+      "step": 13248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9044815301895142,
+      "learning_rate": 4.650039651070579e-05,
+      "loss": 1.3674,
+      "step": 13249
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.923750400543213,
+      "learning_rate": 4.6500132170235265e-05,
+      "loss": 1.514,
+      "step": 13250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6242003440856934,
+      "learning_rate": 4.649986782976474e-05,
+      "loss": 2.632,
+      "step": 13251
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5774638652801514,
+      "learning_rate": 4.649960348929421e-05,
+      "loss": 1.2469,
+      "step": 13252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8054494857788086,
+      "learning_rate": 4.649933914882369e-05,
+      "loss": 1.4258,
+      "step": 13253
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.600506067276001,
+      "learning_rate": 4.649907480835316e-05,
+      "loss": 1.4669,
+      "step": 13254
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6139150857925415,
+      "learning_rate": 4.649881046788263e-05,
+      "loss": 1.3855,
+      "step": 13255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2779741287231445,
+      "learning_rate": 4.649854612741211e-05,
+      "loss": 1.5344,
+      "step": 13256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4479587078094482,
+      "learning_rate": 4.649828178694158e-05,
+      "loss": 1.2412,
+      "step": 13257
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6087925434112549,
+      "learning_rate": 4.649801744647106e-05,
+      "loss": 1.2553,
+      "step": 13258
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7869395017623901,
+      "learning_rate": 4.649775310600053e-05,
+      "loss": 1.4118,
+      "step": 13259
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4449485540390015,
+      "learning_rate": 4.649748876553001e-05,
+      "loss": 1.2326,
+      "step": 13260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9394539594650269,
+      "learning_rate": 4.6497224425059475e-05,
+      "loss": 1.3312,
+      "step": 13261
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.098771095275879,
+      "learning_rate": 4.6496960084588955e-05,
+      "loss": 1.4184,
+      "step": 13262
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.894965887069702,
+      "learning_rate": 4.649669574411843e-05,
+      "loss": 1.5335,
+      "step": 13263
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8073571920394897,
+      "learning_rate": 4.64964314036479e-05,
+      "loss": 1.398,
+      "step": 13264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.767932653427124,
+      "learning_rate": 4.6496167063177376e-05,
+      "loss": 1.5909,
+      "step": 13265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6768927574157715,
+      "learning_rate": 4.649590272270685e-05,
+      "loss": 1.387,
+      "step": 13266
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3898662328720093,
+      "learning_rate": 4.649563838223632e-05,
+      "loss": 1.3574,
+      "step": 13267
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6268908977508545,
+      "learning_rate": 4.6495374041765796e-05,
+      "loss": 1.4623,
+      "step": 13268
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.41674542427063,
+      "learning_rate": 4.649510970129527e-05,
+      "loss": 1.6659,
+      "step": 13269
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5822536945343018,
+      "learning_rate": 4.6494845360824743e-05,
+      "loss": 1.6396,
+      "step": 13270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7156028747558594,
+      "learning_rate": 4.649458102035422e-05,
+      "loss": 1.5053,
+      "step": 13271
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6738641262054443,
+      "learning_rate": 4.649431667988369e-05,
+      "loss": 1.3196,
+      "step": 13272
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2865099906921387,
+      "learning_rate": 4.649405233941317e-05,
+      "loss": 1.2839,
+      "step": 13273
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8071880340576172,
+      "learning_rate": 4.649378799894264e-05,
+      "loss": 1.4638,
+      "step": 13274
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7586263418197632,
+      "learning_rate": 4.649352365847212e-05,
+      "loss": 1.3288,
+      "step": 13275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4602699279785156,
+      "learning_rate": 4.6493259318001585e-05,
+      "loss": 1.3946,
+      "step": 13276
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7724361419677734,
+      "learning_rate": 4.6492994977531065e-05,
+      "loss": 1.3431,
+      "step": 13277
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.737189769744873,
+      "learning_rate": 4.649273063706054e-05,
+      "loss": 1.3861,
+      "step": 13278
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.673932671546936,
+      "learning_rate": 4.649246629659001e-05,
+      "loss": 1.359,
+      "step": 13279
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8431682586669922,
+      "learning_rate": 4.6492201956119486e-05,
+      "loss": 1.3473,
+      "step": 13280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8148587942123413,
+      "learning_rate": 4.649193761564896e-05,
+      "loss": 1.4442,
+      "step": 13281
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6338586807250977,
+      "learning_rate": 4.649167327517843e-05,
+      "loss": 1.2954,
+      "step": 13282
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.75819730758667,
+      "learning_rate": 4.649140893470791e-05,
+      "loss": 1.3922,
+      "step": 13283
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6728438138961792,
+      "learning_rate": 4.649114459423738e-05,
+      "loss": 1.3228,
+      "step": 13284
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5550591945648193,
+      "learning_rate": 4.6490880253766854e-05,
+      "loss": 1.3803,
+      "step": 13285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2764447927474976,
+      "learning_rate": 4.649061591329633e-05,
+      "loss": 1.3368,
+      "step": 13286
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6292694807052612,
+      "learning_rate": 4.64903515728258e-05,
+      "loss": 1.3217,
+      "step": 13287
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7752127647399902,
+      "learning_rate": 4.649008723235528e-05,
+      "loss": 1.4909,
+      "step": 13288
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7312994003295898,
+      "learning_rate": 4.648982289188475e-05,
+      "loss": 1.3679,
+      "step": 13289
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4916948080062866,
+      "learning_rate": 4.648955855141423e-05,
+      "loss": 1.3504,
+      "step": 13290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2666449546813965,
+      "learning_rate": 4.6489294210943695e-05,
+      "loss": 1.3944,
+      "step": 13291
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.700852632522583,
+      "learning_rate": 4.648902987047317e-05,
+      "loss": 1.3743,
+      "step": 13292
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.891884446144104,
+      "learning_rate": 4.648876553000265e-05,
+      "loss": 1.4177,
+      "step": 13293
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.110553026199341,
+      "learning_rate": 4.6488501189532116e-05,
+      "loss": 1.3906,
+      "step": 13294
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.964251756668091,
+      "learning_rate": 4.6488236849061596e-05,
+      "loss": 1.2015,
+      "step": 13295
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5762414932250977,
+      "learning_rate": 4.648797250859106e-05,
+      "loss": 1.3528,
+      "step": 13296
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1118967533111572,
+      "learning_rate": 4.6487708168120543e-05,
+      "loss": 1.4263,
+      "step": 13297
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.23978853225708,
+      "learning_rate": 4.648744382765002e-05,
+      "loss": 1.5838,
+      "step": 13298
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6632652282714844,
+      "learning_rate": 4.648717948717949e-05,
+      "loss": 2.1403,
+      "step": 13299
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8203386068344116,
+      "learning_rate": 4.6486915146708964e-05,
+      "loss": 1.3828,
+      "step": 13300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.620293140411377,
+      "learning_rate": 4.648665080623844e-05,
+      "loss": 1.3646,
+      "step": 13301
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6295892000198364,
+      "learning_rate": 4.648638646576791e-05,
+      "loss": 1.3997,
+      "step": 13302
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0639452934265137,
+      "learning_rate": 4.6486122125297385e-05,
+      "loss": 1.5973,
+      "step": 13303
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8242195844650269,
+      "learning_rate": 4.648585778482686e-05,
+      "loss": 1.3946,
+      "step": 13304
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3753001689910889,
+      "learning_rate": 4.648559344435633e-05,
+      "loss": 1.2841,
+      "step": 13305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8370239734649658,
+      "learning_rate": 4.6485329103885806e-05,
+      "loss": 1.4485,
+      "step": 13306
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5794817209243774,
+      "learning_rate": 4.648506476341528e-05,
+      "loss": 1.2754,
+      "step": 13307
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6482759714126587,
+      "learning_rate": 4.648480042294476e-05,
+      "loss": 1.42,
+      "step": 13308
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8078689575195312,
+      "learning_rate": 4.6484536082474226e-05,
+      "loss": 1.5291,
+      "step": 13309
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.477869749069214,
+      "learning_rate": 4.6484271742003707e-05,
+      "loss": 1.544,
+      "step": 13310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.073472499847412,
+      "learning_rate": 4.6484007401533173e-05,
+      "loss": 1.5055,
+      "step": 13311
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0267202854156494,
+      "learning_rate": 4.6483743061062654e-05,
+      "loss": 1.4493,
+      "step": 13312
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9307307004928589,
+      "learning_rate": 4.648347872059213e-05,
+      "loss": 1.4636,
+      "step": 13313
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.23816180229187,
+      "learning_rate": 4.64832143801216e-05,
+      "loss": 1.2481,
+      "step": 13314
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5206735134124756,
+      "learning_rate": 4.6482950039651074e-05,
+      "loss": 1.2636,
+      "step": 13315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2156500816345215,
+      "learning_rate": 4.648268569918055e-05,
+      "loss": 1.4305,
+      "step": 13316
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5343759059906006,
+      "learning_rate": 4.648242135871002e-05,
+      "loss": 1.2329,
+      "step": 13317
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.446176528930664,
+      "learning_rate": 4.6482157018239495e-05,
+      "loss": 1.5229,
+      "step": 13318
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.81031334400177,
+      "learning_rate": 4.648189267776897e-05,
+      "loss": 1.4681,
+      "step": 13319
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5131579637527466,
+      "learning_rate": 4.648162833729844e-05,
+      "loss": 1.3678,
+      "step": 13320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.16375994682312,
+      "learning_rate": 4.6481363996827916e-05,
+      "loss": 1.4085,
+      "step": 13321
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8195894956588745,
+      "learning_rate": 4.648109965635739e-05,
+      "loss": 1.4446,
+      "step": 13322
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1395301818847656,
+      "learning_rate": 4.648083531588687e-05,
+      "loss": 1.3749,
+      "step": 13323
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5791475772857666,
+      "learning_rate": 4.6480570975416337e-05,
+      "loss": 1.3164,
+      "step": 13324
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9407912492752075,
+      "learning_rate": 4.648030663494582e-05,
+      "loss": 1.3799,
+      "step": 13325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9230027198791504,
+      "learning_rate": 4.6480042294475284e-05,
+      "loss": 1.3876,
+      "step": 13326
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6338715553283691,
+      "learning_rate": 4.647977795400476e-05,
+      "loss": 1.4904,
+      "step": 13327
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7268576622009277,
+      "learning_rate": 4.647951361353424e-05,
+      "loss": 1.5319,
+      "step": 13328
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7471662759780884,
+      "learning_rate": 4.6479249273063704e-05,
+      "loss": 1.4373,
+      "step": 13329
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2912077903747559,
+      "learning_rate": 4.6478984932593185e-05,
+      "loss": 1.1946,
+      "step": 13330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.931012749671936,
+      "learning_rate": 4.647872059212265e-05,
+      "loss": 1.4596,
+      "step": 13331
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4664441347122192,
+      "learning_rate": 4.647845625165213e-05,
+      "loss": 1.1235,
+      "step": 13332
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4834551811218262,
+      "learning_rate": 4.6478191911181606e-05,
+      "loss": 1.3837,
+      "step": 13333
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6197779178619385,
+      "learning_rate": 4.647792757071108e-05,
+      "loss": 1.3766,
+      "step": 13334
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3775203227996826,
+      "learning_rate": 4.647766323024055e-05,
+      "loss": 1.3269,
+      "step": 13335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8176751136779785,
+      "learning_rate": 4.6477398889770026e-05,
+      "loss": 1.401,
+      "step": 13336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9938007593154907,
+      "learning_rate": 4.64771345492995e-05,
+      "loss": 1.4724,
+      "step": 13337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8338770866394043,
+      "learning_rate": 4.647687020882897e-05,
+      "loss": 1.4637,
+      "step": 13338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.108992576599121,
+      "learning_rate": 4.647660586835845e-05,
+      "loss": 1.3925,
+      "step": 13339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.79927396774292,
+      "learning_rate": 4.647634152788792e-05,
+      "loss": 1.3845,
+      "step": 13340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1000566482543945,
+      "learning_rate": 4.6476077187417394e-05,
+      "loss": 1.4542,
+      "step": 13341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8157451152801514,
+      "learning_rate": 4.647581284694687e-05,
+      "loss": 1.451,
+      "step": 13342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5470153093338013,
+      "learning_rate": 4.647554850647635e-05,
+      "loss": 1.4283,
+      "step": 13343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4067370891571045,
+      "learning_rate": 4.6475284166005815e-05,
+      "loss": 1.3509,
+      "step": 13344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8440887928009033,
+      "learning_rate": 4.6475019825535295e-05,
+      "loss": 1.3227,
+      "step": 13345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.056931495666504,
+      "learning_rate": 4.647475548506476e-05,
+      "loss": 1.4116,
+      "step": 13346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.658430814743042,
+      "learning_rate": 4.647449114459424e-05,
+      "loss": 1.3612,
+      "step": 13347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.292569875717163,
+      "learning_rate": 4.6474226804123716e-05,
+      "loss": 1.5672,
+      "step": 13348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7793991565704346,
+      "learning_rate": 4.647396246365319e-05,
+      "loss": 1.3104,
+      "step": 13349
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6950596570968628,
+      "learning_rate": 4.647369812318266e-05,
+      "loss": 1.3364,
+      "step": 13350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.478074312210083,
+      "learning_rate": 4.6473433782712137e-05,
+      "loss": 1.276,
+      "step": 13351
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.420123815536499,
+      "learning_rate": 4.647316944224161e-05,
+      "loss": 1.1727,
+      "step": 13352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5873939990997314,
+      "learning_rate": 4.6472905101771084e-05,
+      "loss": 1.5724,
+      "step": 13353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.488867998123169,
+      "learning_rate": 4.647264076130056e-05,
+      "loss": 1.3628,
+      "step": 13354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9338449239730835,
+      "learning_rate": 4.647237642083003e-05,
+      "loss": 1.4115,
+      "step": 13355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.56207013130188,
+      "learning_rate": 4.6472112080359504e-05,
+      "loss": 1.4171,
+      "step": 13356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8983709812164307,
+      "learning_rate": 4.647184773988898e-05,
+      "loss": 1.3744,
+      "step": 13357
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.409085273742676,
+      "learning_rate": 4.647158339941846e-05,
+      "loss": 1.5043,
+      "step": 13358
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6529541015625,
+      "learning_rate": 4.6471319058947925e-05,
+      "loss": 1.2939,
+      "step": 13359
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2551348209381104,
+      "learning_rate": 4.6471054718477405e-05,
+      "loss": 1.3668,
+      "step": 13360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8994840383529663,
+      "learning_rate": 4.647079037800687e-05,
+      "loss": 1.4757,
+      "step": 13361
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3890013694763184,
+      "learning_rate": 4.647052603753635e-05,
+      "loss": 2.3805,
+      "step": 13362
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5199517011642456,
+      "learning_rate": 4.6470261697065826e-05,
+      "loss": 1.4291,
+      "step": 13363
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9236924648284912,
+      "learning_rate": 4.646999735659529e-05,
+      "loss": 1.4581,
+      "step": 13364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0595948696136475,
+      "learning_rate": 4.646973301612477e-05,
+      "loss": 1.5386,
+      "step": 13365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.856735110282898,
+      "learning_rate": 4.646946867565424e-05,
+      "loss": 1.4177,
+      "step": 13366
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9954642057418823,
+      "learning_rate": 4.646920433518372e-05,
+      "loss": 1.5102,
+      "step": 13367
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8029669523239136,
+      "learning_rate": 4.6468939994713194e-05,
+      "loss": 1.3402,
+      "step": 13368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5215178728103638,
+      "learning_rate": 4.646867565424267e-05,
+      "loss": 1.2836,
+      "step": 13369
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7652487754821777,
+      "learning_rate": 4.646841131377214e-05,
+      "loss": 1.3835,
+      "step": 13370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5165903568267822,
+      "learning_rate": 4.6468146973301615e-05,
+      "loss": 1.3296,
+      "step": 13371
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9629287719726562,
+      "learning_rate": 4.646788263283109e-05,
+      "loss": 1.4909,
+      "step": 13372
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.593906283378601,
+      "learning_rate": 4.646761829236056e-05,
+      "loss": 1.3595,
+      "step": 13373
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7627607583999634,
+      "learning_rate": 4.6467353951890035e-05,
+      "loss": 1.3489,
+      "step": 13374
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9658470153808594,
+      "learning_rate": 4.646708961141951e-05,
+      "loss": 1.2824,
+      "step": 13375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6566561460494995,
+      "learning_rate": 4.646682527094898e-05,
+      "loss": 1.5161,
+      "step": 13376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3229668140411377,
+      "learning_rate": 4.6466560930478456e-05,
+      "loss": 1.2165,
+      "step": 13377
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0259437561035156,
+      "learning_rate": 4.6466296590007937e-05,
+      "loss": 1.2435,
+      "step": 13378
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.681882858276367,
+      "learning_rate": 4.64660322495374e-05,
+      "loss": 1.7178,
+      "step": 13379
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1567468643188477,
+      "learning_rate": 4.6465767909066884e-05,
+      "loss": 1.8104,
+      "step": 13380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8534314632415771,
+      "learning_rate": 4.646550356859635e-05,
+      "loss": 1.4587,
+      "step": 13381
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.080465316772461,
+      "learning_rate": 4.646523922812583e-05,
+      "loss": 1.399,
+      "step": 13382
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7030872106552124,
+      "learning_rate": 4.6464974887655304e-05,
+      "loss": 1.4775,
+      "step": 13383
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8526349067687988,
+      "learning_rate": 4.646471054718478e-05,
+      "loss": 1.302,
+      "step": 13384
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0611836910247803,
+      "learning_rate": 4.646444620671425e-05,
+      "loss": 1.3447,
+      "step": 13385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.480500340461731,
+      "learning_rate": 4.6464181866243725e-05,
+      "loss": 1.422,
+      "step": 13386
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.413118600845337,
+      "learning_rate": 4.64639175257732e-05,
+      "loss": 1.3114,
+      "step": 13387
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.004183530807495,
+      "learning_rate": 4.646365318530267e-05,
+      "loss": 1.367,
+      "step": 13388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5676891803741455,
+      "learning_rate": 4.6463388844832146e-05,
+      "loss": 1.3146,
+      "step": 13389
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5405247211456299,
+      "learning_rate": 4.646312450436162e-05,
+      "loss": 1.358,
+      "step": 13390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8044476509094238,
+      "learning_rate": 4.646286016389109e-05,
+      "loss": 1.4273,
+      "step": 13391
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8425617218017578,
+      "learning_rate": 4.6462595823420567e-05,
+      "loss": 1.4896,
+      "step": 13392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8039215803146362,
+      "learning_rate": 4.646233148295005e-05,
+      "loss": 1.3243,
+      "step": 13393
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5600614547729492,
+      "learning_rate": 4.6462067142479514e-05,
+      "loss": 1.3707,
+      "step": 13394
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.498180866241455,
+      "learning_rate": 4.6461802802008994e-05,
+      "loss": 1.46,
+      "step": 13395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5215013027191162,
+      "learning_rate": 4.646153846153846e-05,
+      "loss": 1.1551,
+      "step": 13396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.709798812866211,
+      "learning_rate": 4.646127412106794e-05,
+      "loss": 1.3843,
+      "step": 13397
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.090538501739502,
+      "learning_rate": 4.6461009780597415e-05,
+      "loss": 1.4516,
+      "step": 13398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.06990647315979,
+      "learning_rate": 4.646074544012688e-05,
+      "loss": 1.2442,
+      "step": 13399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7843163013458252,
+      "learning_rate": 4.646048109965636e-05,
+      "loss": 1.3047,
+      "step": 13400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6543562412261963,
+      "learning_rate": 4.646021675918583e-05,
+      "loss": 1.2863,
+      "step": 13401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8976150751113892,
+      "learning_rate": 4.645995241871531e-05,
+      "loss": 1.3409,
+      "step": 13402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3790860176086426,
+      "learning_rate": 4.645968807824478e-05,
+      "loss": 1.4159,
+      "step": 13403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8436412811279297,
+      "learning_rate": 4.6459423737774256e-05,
+      "loss": 1.3884,
+      "step": 13404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3484007120132446,
+      "learning_rate": 4.645915939730373e-05,
+      "loss": 1.2785,
+      "step": 13405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.583093523979187,
+      "learning_rate": 4.64588950568332e-05,
+      "loss": 1.3943,
+      "step": 13406
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5647284984588623,
+      "learning_rate": 4.645863071636268e-05,
+      "loss": 1.3617,
+      "step": 13407
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9249773025512695,
+      "learning_rate": 4.645836637589216e-05,
+      "loss": 1.3887,
+      "step": 13408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5522105693817139,
+      "learning_rate": 4.6458102035421624e-05,
+      "loss": 1.268,
+      "step": 13409
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6681166887283325,
+      "learning_rate": 4.64578376949511e-05,
+      "loss": 1.308,
+      "step": 13410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6989507675170898,
+      "learning_rate": 4.645757335448057e-05,
+      "loss": 1.3229,
+      "step": 13411
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.802815318107605,
+      "learning_rate": 4.6457309014010045e-05,
+      "loss": 1.3349,
+      "step": 13412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7170380353927612,
+      "learning_rate": 4.6457044673539525e-05,
+      "loss": 1.547,
+      "step": 13413
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.537226915359497,
+      "learning_rate": 4.645678033306899e-05,
+      "loss": 1.5433,
+      "step": 13414
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9452379941940308,
+      "learning_rate": 4.645651599259847e-05,
+      "loss": 1.3665,
+      "step": 13415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5040042400360107,
+      "learning_rate": 4.645625165212794e-05,
+      "loss": 1.358,
+      "step": 13416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.238067626953125,
+      "learning_rate": 4.645598731165742e-05,
+      "loss": 1.562,
+      "step": 13417
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4211175441741943,
+      "learning_rate": 4.645572297118689e-05,
+      "loss": 1.3051,
+      "step": 13418
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8351092338562012,
+      "learning_rate": 4.6455458630716366e-05,
+      "loss": 2.4729,
+      "step": 13419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7436586618423462,
+      "learning_rate": 4.645519429024584e-05,
+      "loss": 1.344,
+      "step": 13420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6021270751953125,
+      "learning_rate": 4.6454929949775314e-05,
+      "loss": 1.3931,
+      "step": 13421
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1189756393432617,
+      "learning_rate": 4.645466560930479e-05,
+      "loss": 1.5843,
+      "step": 13422
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9249424934387207,
+      "learning_rate": 4.645440126883426e-05,
+      "loss": 1.5242,
+      "step": 13423
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.943450927734375,
+      "learning_rate": 4.6454136928363734e-05,
+      "loss": 1.4206,
+      "step": 13424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.072004556655884,
+      "learning_rate": 4.645387258789321e-05,
+      "loss": 1.5289,
+      "step": 13425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2435998916625977,
+      "learning_rate": 4.645360824742268e-05,
+      "loss": 2.8552,
+      "step": 13426
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6086903810501099,
+      "learning_rate": 4.6453343906952155e-05,
+      "loss": 1.4282,
+      "step": 13427
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8023278713226318,
+      "learning_rate": 4.6453079566481635e-05,
+      "loss": 1.3873,
+      "step": 13428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8599321842193604,
+      "learning_rate": 4.64528152260111e-05,
+      "loss": 1.3861,
+      "step": 13429
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.647821307182312,
+      "learning_rate": 4.645255088554058e-05,
+      "loss": 1.3527,
+      "step": 13430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4784482717514038,
+      "learning_rate": 4.645228654507005e-05,
+      "loss": 1.2461,
+      "step": 13431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4886927604675293,
+      "learning_rate": 4.645202220459953e-05,
+      "loss": 1.4163,
+      "step": 13432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.639392614364624,
+      "learning_rate": 4.6451757864129e-05,
+      "loss": 1.2969,
+      "step": 13433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9462319612503052,
+      "learning_rate": 4.645149352365848e-05,
+      "loss": 1.3786,
+      "step": 13434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9217292070388794,
+      "learning_rate": 4.645122918318795e-05,
+      "loss": 2.0385,
+      "step": 13435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5866485834121704,
+      "learning_rate": 4.645096484271742e-05,
+      "loss": 1.4563,
+      "step": 13436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5418970584869385,
+      "learning_rate": 4.64507005022469e-05,
+      "loss": 1.4564,
+      "step": 13437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7697595357894897,
+      "learning_rate": 4.645043616177637e-05,
+      "loss": 1.3238,
+      "step": 13438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5387136936187744,
+      "learning_rate": 4.6450171821305845e-05,
+      "loss": 1.2799,
+      "step": 13439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.611266851425171,
+      "learning_rate": 4.644990748083532e-05,
+      "loss": 1.4189,
+      "step": 13440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.874930500984192,
+      "learning_rate": 4.644964314036479e-05,
+      "loss": 1.5637,
+      "step": 13441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5268275737762451,
+      "learning_rate": 4.6449378799894265e-05,
+      "loss": 1.3212,
+      "step": 13442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4201480150222778,
+      "learning_rate": 4.6449114459423746e-05,
+      "loss": 1.2714,
+      "step": 13443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2963216304779053,
+      "learning_rate": 4.644885011895321e-05,
+      "loss": 1.4368,
+      "step": 13444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6222758293151855,
+      "learning_rate": 4.6448585778482686e-05,
+      "loss": 1.2519,
+      "step": 13445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.394199848175049,
+      "learning_rate": 4.644832143801216e-05,
+      "loss": 1.502,
+      "step": 13446
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.888434886932373,
+      "learning_rate": 4.644805709754163e-05,
+      "loss": 2.7882,
+      "step": 13447
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3255853652954102,
+      "learning_rate": 4.6447792757071114e-05,
+      "loss": 1.2453,
+      "step": 13448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8072808980941772,
+      "learning_rate": 4.644752841660058e-05,
+      "loss": 1.2565,
+      "step": 13449
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.235938549041748,
+      "learning_rate": 4.644726407613006e-05,
+      "loss": 1.2612,
+      "step": 13450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7538832426071167,
+      "learning_rate": 4.644699973565953e-05,
+      "loss": 1.492,
+      "step": 13451
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4769957065582275,
+      "learning_rate": 4.644673539518901e-05,
+      "loss": 1.4694,
+      "step": 13452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5100414752960205,
+      "learning_rate": 4.644647105471848e-05,
+      "loss": 1.3658,
+      "step": 13453
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7456445693969727,
+      "learning_rate": 4.6446206714247955e-05,
+      "loss": 1.4108,
+      "step": 13454
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.153010845184326,
+      "learning_rate": 4.644594237377743e-05,
+      "loss": 1.4575,
+      "step": 13455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9718557596206665,
+      "learning_rate": 4.64456780333069e-05,
+      "loss": 1.4996,
+      "step": 13456
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0949599742889404,
+      "learning_rate": 4.6445413692836376e-05,
+      "loss": 1.6499,
+      "step": 13457
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7344502210617065,
+      "learning_rate": 4.644514935236585e-05,
+      "loss": 1.2517,
+      "step": 13458
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5778485536575317,
+      "learning_rate": 4.644488501189532e-05,
+      "loss": 1.3408,
+      "step": 13459
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.539617657661438,
+      "learning_rate": 4.6444620671424796e-05,
+      "loss": 1.6955,
+      "step": 13460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7702182531356812,
+      "learning_rate": 4.644435633095427e-05,
+      "loss": 1.5789,
+      "step": 13461
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.646950125694275,
+      "learning_rate": 4.6444091990483744e-05,
+      "loss": 1.4038,
+      "step": 13462
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0381088256835938,
+      "learning_rate": 4.6443827650013224e-05,
+      "loss": 1.5644,
+      "step": 13463
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2223896980285645,
+      "learning_rate": 4.644356330954269e-05,
+      "loss": 1.5842,
+      "step": 13464
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9385055303573608,
+      "learning_rate": 4.644329896907217e-05,
+      "loss": 1.3631,
+      "step": 13465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.974365234375,
+      "learning_rate": 4.644303462860164e-05,
+      "loss": 1.3998,
+      "step": 13466
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5698963403701782,
+      "learning_rate": 4.644277028813112e-05,
+      "loss": 1.2997,
+      "step": 13467
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0898709297180176,
+      "learning_rate": 4.644250594766059e-05,
+      "loss": 1.7048,
+      "step": 13468
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8059824705123901,
+      "learning_rate": 4.6442241607190065e-05,
+      "loss": 1.2847,
+      "step": 13469
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.718356728553772,
+      "learning_rate": 4.644197726671954e-05,
+      "loss": 1.4125,
+      "step": 13470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5364099740982056,
+      "learning_rate": 4.644171292624901e-05,
+      "loss": 1.3292,
+      "step": 13471
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4142515659332275,
+      "learning_rate": 4.6441448585778486e-05,
+      "loss": 1.3853,
+      "step": 13472
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7167125940322876,
+      "learning_rate": 4.644118424530796e-05,
+      "loss": 1.2823,
+      "step": 13473
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7930963039398193,
+      "learning_rate": 4.644091990483743e-05,
+      "loss": 1.5255,
+      "step": 13474
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.767439603805542,
+      "learning_rate": 4.644065556436691e-05,
+      "loss": 1.3352,
+      "step": 13475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5660120248794556,
+      "learning_rate": 4.644039122389638e-05,
+      "loss": 1.3742,
+      "step": 13476
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6015801429748535,
+      "learning_rate": 4.6440126883425854e-05,
+      "loss": 1.4171,
+      "step": 13477
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.725083351135254,
+      "learning_rate": 4.6439862542955334e-05,
+      "loss": 1.5331,
+      "step": 13478
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.81541109085083,
+      "learning_rate": 4.64395982024848e-05,
+      "loss": 1.4871,
+      "step": 13479
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9313780069351196,
+      "learning_rate": 4.643933386201428e-05,
+      "loss": 1.5368,
+      "step": 13480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5929176807403564,
+      "learning_rate": 4.643906952154375e-05,
+      "loss": 1.3871,
+      "step": 13481
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3799902200698853,
+      "learning_rate": 4.643880518107322e-05,
+      "loss": 1.4675,
+      "step": 13482
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6383881568908691,
+      "learning_rate": 4.64385408406027e-05,
+      "loss": 1.4306,
+      "step": 13483
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.766343355178833,
+      "learning_rate": 4.643827650013217e-05,
+      "loss": 1.3852,
+      "step": 13484
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7771899700164795,
+      "learning_rate": 4.643801215966165e-05,
+      "loss": 1.4118,
+      "step": 13485
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6025238037109375,
+      "learning_rate": 4.6437747819191116e-05,
+      "loss": 1.3755,
+      "step": 13486
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.667283058166504,
+      "learning_rate": 4.6437483478720596e-05,
+      "loss": 1.1935,
+      "step": 13487
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4782817363739014,
+      "learning_rate": 4.643721913825007e-05,
+      "loss": 1.2446,
+      "step": 13488
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0642552375793457,
+      "learning_rate": 4.6436954797779544e-05,
+      "loss": 1.4541,
+      "step": 13489
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6665512323379517,
+      "learning_rate": 4.643669045730902e-05,
+      "loss": 1.3404,
+      "step": 13490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.694216012954712,
+      "learning_rate": 4.643642611683849e-05,
+      "loss": 2.4279,
+      "step": 13491
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9141671657562256,
+      "learning_rate": 4.6436161776367964e-05,
+      "loss": 1.2088,
+      "step": 13492
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8679454326629639,
+      "learning_rate": 4.643589743589744e-05,
+      "loss": 1.4676,
+      "step": 13493
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9082626104354858,
+      "learning_rate": 4.643563309542691e-05,
+      "loss": 1.2759,
+      "step": 13494
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.349703311920166,
+      "learning_rate": 4.6435368754956385e-05,
+      "loss": 1.7652,
+      "step": 13495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.19572377204895,
+      "learning_rate": 4.643510441448586e-05,
+      "loss": 1.7212,
+      "step": 13496
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.725062608718872,
+      "learning_rate": 4.643484007401533e-05,
+      "loss": 1.329,
+      "step": 13497
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8124014139175415,
+      "learning_rate": 4.643457573354481e-05,
+      "loss": 1.5508,
+      "step": 13498
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6651474237442017,
+      "learning_rate": 4.643431139307428e-05,
+      "loss": 1.3542,
+      "step": 13499
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.555023193359375,
+      "learning_rate": 4.643404705260376e-05,
+      "loss": 1.1684,
+      "step": 13500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.873701810836792,
+      "learning_rate": 4.6433782712133226e-05,
+      "loss": 1.2696,
+      "step": 13501
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6365723609924316,
+      "learning_rate": 4.643351837166271e-05,
+      "loss": 1.3506,
+      "step": 13502
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.684842348098755,
+      "learning_rate": 4.643325403119218e-05,
+      "loss": 1.4503,
+      "step": 13503
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7794485092163086,
+      "learning_rate": 4.6432989690721654e-05,
+      "loss": 1.3667,
+      "step": 13504
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2904400825500488,
+      "learning_rate": 4.643272535025113e-05,
+      "loss": 1.0402,
+      "step": 13505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8260078430175781,
+      "learning_rate": 4.64324610097806e-05,
+      "loss": 1.3833,
+      "step": 13506
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9967002868652344,
+      "learning_rate": 4.6432196669310075e-05,
+      "loss": 1.4442,
+      "step": 13507
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1196045875549316,
+      "learning_rate": 4.643193232883955e-05,
+      "loss": 1.4925,
+      "step": 13508
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.538891315460205,
+      "learning_rate": 4.643166798836902e-05,
+      "loss": 1.4399,
+      "step": 13509
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.410524845123291,
+      "learning_rate": 4.6431403647898495e-05,
+      "loss": 1.2313,
+      "step": 13510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.358670949935913,
+      "learning_rate": 4.643113930742797e-05,
+      "loss": 1.4295,
+      "step": 13511
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9057039022445679,
+      "learning_rate": 4.643087496695744e-05,
+      "loss": 1.4166,
+      "step": 13512
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7645796537399292,
+      "learning_rate": 4.643061062648692e-05,
+      "loss": 1.3915,
+      "step": 13513
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5301988124847412,
+      "learning_rate": 4.643034628601639e-05,
+      "loss": 1.29,
+      "step": 13514
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.274434804916382,
+      "learning_rate": 4.643008194554587e-05,
+      "loss": 1.5299,
+      "step": 13515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7439684867858887,
+      "learning_rate": 4.642981760507534e-05,
+      "loss": 1.4555,
+      "step": 13516
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.017204523086548,
+      "learning_rate": 4.642955326460481e-05,
+      "loss": 1.3671,
+      "step": 13517
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9415559768676758,
+      "learning_rate": 4.642928892413429e-05,
+      "loss": 1.4529,
+      "step": 13518
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.685423493385315,
+      "learning_rate": 4.642902458366376e-05,
+      "loss": 1.407,
+      "step": 13519
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5956578254699707,
+      "learning_rate": 4.642876024319324e-05,
+      "loss": 1.5716,
+      "step": 13520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6994657516479492,
+      "learning_rate": 4.6428495902722705e-05,
+      "loss": 1.408,
+      "step": 13521
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6941609382629395,
+      "learning_rate": 4.6428231562252185e-05,
+      "loss": 1.315,
+      "step": 13522
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3048906326293945,
+      "learning_rate": 4.642796722178166e-05,
+      "loss": 1.4201,
+      "step": 13523
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.735978603363037,
+      "learning_rate": 4.642770288131113e-05,
+      "loss": 1.3558,
+      "step": 13524
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9231247901916504,
+      "learning_rate": 4.6427438540840606e-05,
+      "loss": 1.427,
+      "step": 13525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6223260164260864,
+      "learning_rate": 4.642717420037008e-05,
+      "loss": 1.3528,
+      "step": 13526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7838472127914429,
+      "learning_rate": 4.642690985989955e-05,
+      "loss": 1.4839,
+      "step": 13527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6796125173568726,
+      "learning_rate": 4.6426645519429026e-05,
+      "loss": 1.4387,
+      "step": 13528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6406735181808472,
+      "learning_rate": 4.64263811789585e-05,
+      "loss": 1.3653,
+      "step": 13529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3798061609268188,
+      "learning_rate": 4.6426116838487974e-05,
+      "loss": 1.1964,
+      "step": 13530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9825485944747925,
+      "learning_rate": 4.642585249801745e-05,
+      "loss": 1.4775,
+      "step": 13531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5993211269378662,
+      "learning_rate": 4.642558815754692e-05,
+      "loss": 1.2965,
+      "step": 13532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6143007278442383,
+      "learning_rate": 4.64253238170764e-05,
+      "loss": 1.3751,
+      "step": 13533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9653904438018799,
+      "learning_rate": 4.642505947660587e-05,
+      "loss": 1.5321,
+      "step": 13534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0564651489257812,
+      "learning_rate": 4.642479513613535e-05,
+      "loss": 1.4053,
+      "step": 13535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.908205270767212,
+      "learning_rate": 4.6424530795664815e-05,
+      "loss": 1.4465,
+      "step": 13536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2288708686828613,
+      "learning_rate": 4.6424266455194295e-05,
+      "loss": 1.4371,
+      "step": 13537
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1945128440856934,
+      "learning_rate": 4.642400211472377e-05,
+      "loss": 1.5301,
+      "step": 13538
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.122542381286621,
+      "learning_rate": 4.642373777425324e-05,
+      "loss": 1.5452,
+      "step": 13539
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.499557614326477,
+      "learning_rate": 4.6423473433782716e-05,
+      "loss": 1.381,
+      "step": 13540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6071540117263794,
+      "learning_rate": 4.642320909331219e-05,
+      "loss": 1.3648,
+      "step": 13541
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4365360736846924,
+      "learning_rate": 4.642294475284166e-05,
+      "loss": 1.4506,
+      "step": 13542
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.139002799987793,
+      "learning_rate": 4.642268041237114e-05,
+      "loss": 1.3319,
+      "step": 13543
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5986884832382202,
+      "learning_rate": 4.642241607190061e-05,
+      "loss": 1.3876,
+      "step": 13544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6245734691619873,
+      "learning_rate": 4.6422151731430084e-05,
+      "loss": 1.3105,
+      "step": 13545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3639212846755981,
+      "learning_rate": 4.642188739095956e-05,
+      "loss": 1.2494,
+      "step": 13546
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6616822481155396,
+      "learning_rate": 4.642162305048903e-05,
+      "loss": 1.3627,
+      "step": 13547
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9093698263168335,
+      "learning_rate": 4.642135871001851e-05,
+      "loss": 1.3507,
+      "step": 13548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.246476173400879,
+      "learning_rate": 4.642109436954798e-05,
+      "loss": 1.5365,
+      "step": 13549
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7685528993606567,
+      "learning_rate": 4.642083002907746e-05,
+      "loss": 1.2656,
+      "step": 13550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8109867572784424,
+      "learning_rate": 4.6420565688606925e-05,
+      "loss": 1.4245,
+      "step": 13551
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5679574012756348,
+      "learning_rate": 4.6420301348136406e-05,
+      "loss": 1.3264,
+      "step": 13552
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6689592599868774,
+      "learning_rate": 4.642003700766588e-05,
+      "loss": 1.1809,
+      "step": 13553
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0144217014312744,
+      "learning_rate": 4.6419772667195346e-05,
+      "loss": 1.4836,
+      "step": 13554
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1338772773742676,
+      "learning_rate": 4.6419508326724826e-05,
+      "loss": 1.5161,
+      "step": 13555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9038450717926025,
+      "learning_rate": 4.641924398625429e-05,
+      "loss": 1.4964,
+      "step": 13556
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1119604110717773,
+      "learning_rate": 4.6418979645783773e-05,
+      "loss": 1.5455,
+      "step": 13557
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109010696411133,
+      "learning_rate": 4.641871530531325e-05,
+      "loss": 1.5669,
+      "step": 13558
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3340983390808105,
+      "learning_rate": 4.641845096484272e-05,
+      "loss": 1.5311,
+      "step": 13559
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4984140396118164,
+      "learning_rate": 4.6418186624372194e-05,
+      "loss": 1.2564,
+      "step": 13560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.994117021560669,
+      "learning_rate": 4.641792228390167e-05,
+      "loss": 1.2839,
+      "step": 13561
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4890543222427368,
+      "learning_rate": 4.641765794343114e-05,
+      "loss": 1.1843,
+      "step": 13562
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.349792003631592,
+      "learning_rate": 4.6417393602960615e-05,
+      "loss": 1.4933,
+      "step": 13563
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8880929946899414,
+      "learning_rate": 4.641712926249009e-05,
+      "loss": 1.4354,
+      "step": 13564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6324141025543213,
+      "learning_rate": 4.641686492201956e-05,
+      "loss": 1.3293,
+      "step": 13565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9629746675491333,
+      "learning_rate": 4.6416600581549036e-05,
+      "loss": 1.5216,
+      "step": 13566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6474523544311523,
+      "learning_rate": 4.641633624107851e-05,
+      "loss": 1.4456,
+      "step": 13567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.921963095664978,
+      "learning_rate": 4.641607190060799e-05,
+      "loss": 1.5281,
+      "step": 13568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.102900743484497,
+      "learning_rate": 4.6415807560137456e-05,
+      "loss": 1.3808,
+      "step": 13569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6786584854125977,
+      "learning_rate": 4.641554321966694e-05,
+      "loss": 1.2925,
+      "step": 13570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4083988666534424,
+      "learning_rate": 4.6415278879196403e-05,
+      "loss": 1.3757,
+      "step": 13571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.397162675857544,
+      "learning_rate": 4.6415014538725884e-05,
+      "loss": 1.615,
+      "step": 13572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9288501739501953,
+      "learning_rate": 4.641475019825536e-05,
+      "loss": 1.4157,
+      "step": 13573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1394317150115967,
+      "learning_rate": 4.641448585778483e-05,
+      "loss": 1.4042,
+      "step": 13574
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.198084592819214,
+      "learning_rate": 4.6414221517314305e-05,
+      "loss": 1.5105,
+      "step": 13575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7189253568649292,
+      "learning_rate": 4.641395717684378e-05,
+      "loss": 1.3633,
+      "step": 13576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.089456081390381,
+      "learning_rate": 4.641369283637325e-05,
+      "loss": 1.4814,
+      "step": 13577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6142568588256836,
+      "learning_rate": 4.6413428495902725e-05,
+      "loss": 1.6942,
+      "step": 13578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4346346855163574,
+      "learning_rate": 4.64131641554322e-05,
+      "loss": 1.3727,
+      "step": 13579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.84732985496521,
+      "learning_rate": 4.641289981496167e-05,
+      "loss": 1.637,
+      "step": 13580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.50617516040802,
+      "learning_rate": 4.6412635474491146e-05,
+      "loss": 1.3194,
+      "step": 13581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7021077871322632,
+      "learning_rate": 4.641237113402062e-05,
+      "loss": 2.2469,
+      "step": 13582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7590398788452148,
+      "learning_rate": 4.64121067935501e-05,
+      "loss": 1.3798,
+      "step": 13583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7077516317367554,
+      "learning_rate": 4.641184245307957e-05,
+      "loss": 1.4023,
+      "step": 13584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5920395851135254,
+      "learning_rate": 4.641157811260905e-05,
+      "loss": 1.4167,
+      "step": 13585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.944062352180481,
+      "learning_rate": 4.6411313772138514e-05,
+      "loss": 1.3427,
+      "step": 13586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9061028957366943,
+      "learning_rate": 4.6411049431667994e-05,
+      "loss": 1.4481,
+      "step": 13587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7489573955535889,
+      "learning_rate": 4.641078509119747e-05,
+      "loss": 1.351,
+      "step": 13588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3851563930511475,
+      "learning_rate": 4.641052075072694e-05,
+      "loss": 1.5585,
+      "step": 13589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.703160285949707,
+      "learning_rate": 4.6410256410256415e-05,
+      "loss": 1.402,
+      "step": 13590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7419947385787964,
+      "learning_rate": 4.640999206978588e-05,
+      "loss": 1.3641,
+      "step": 13591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.958823323249817,
+      "learning_rate": 4.640972772931536e-05,
+      "loss": 1.452,
+      "step": 13592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8243825435638428,
+      "learning_rate": 4.6409463388844836e-05,
+      "loss": 1.5022,
+      "step": 13593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4874610900878906,
+      "learning_rate": 4.640919904837431e-05,
+      "loss": 1.362,
+      "step": 13594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2546277046203613,
+      "learning_rate": 4.640893470790378e-05,
+      "loss": 1.3318,
+      "step": 13595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7825206518173218,
+      "learning_rate": 4.6408670367433256e-05,
+      "loss": 1.3988,
+      "step": 13596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.052119731903076,
+      "learning_rate": 4.640840602696273e-05,
+      "loss": 1.4825,
+      "step": 13597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4945909976959229,
+      "learning_rate": 4.640814168649221e-05,
+      "loss": 1.3448,
+      "step": 13598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.386582851409912,
+      "learning_rate": 4.640787734602168e-05,
+      "loss": 1.4632,
+      "step": 13599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9703062772750854,
+      "learning_rate": 4.640761300555115e-05,
+      "loss": 1.3952,
+      "step": 13600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.559883713722229,
+      "learning_rate": 4.6407348665080624e-05,
+      "loss": 1.3489,
+      "step": 13601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8760164976119995,
+      "learning_rate": 4.64070843246101e-05,
+      "loss": 1.4732,
+      "step": 13602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.4963250160217285,
+      "learning_rate": 4.640681998413958e-05,
+      "loss": 1.3494,
+      "step": 13603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.849900245666504,
+      "learning_rate": 4.6406555643669045e-05,
+      "loss": 1.2951,
+      "step": 13604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6825700998306274,
+      "learning_rate": 4.6406291303198525e-05,
+      "loss": 1.3807,
+      "step": 13605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.047027826309204,
+      "learning_rate": 4.640602696272799e-05,
+      "loss": 1.5998,
+      "step": 13606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.332288980484009,
+      "learning_rate": 4.640576262225747e-05,
+      "loss": 1.4655,
+      "step": 13607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.602111577987671,
+      "learning_rate": 4.640549828178694e-05,
+      "loss": 1.3594,
+      "step": 13608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.64189612865448,
+      "learning_rate": 4.640523394131642e-05,
+      "loss": 1.2582,
+      "step": 13609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1494383811950684,
+      "learning_rate": 4.640496960084589e-05,
+      "loss": 1.4859,
+      "step": 13610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.410869836807251,
+      "learning_rate": 4.640470526037537e-05,
+      "loss": 1.2654,
+      "step": 13611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5963711738586426,
+      "learning_rate": 4.640444091990484e-05,
+      "loss": 1.351,
+      "step": 13612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4049474000930786,
+      "learning_rate": 4.6404176579434314e-05,
+      "loss": 1.2365,
+      "step": 13613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5825722217559814,
+      "learning_rate": 4.640391223896379e-05,
+      "loss": 1.3506,
+      "step": 13614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.304544687271118,
+      "learning_rate": 4.640364789849326e-05,
+      "loss": 1.4304,
+      "step": 13615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4612245559692383,
+      "learning_rate": 4.6403383558022734e-05,
+      "loss": 1.7804,
+      "step": 13616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1901917457580566,
+      "learning_rate": 4.640311921755221e-05,
+      "loss": 1.5155,
+      "step": 13617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0075669288635254,
+      "learning_rate": 4.640285487708168e-05,
+      "loss": 1.2514,
+      "step": 13618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9897245168685913,
+      "learning_rate": 4.6402590536611155e-05,
+      "loss": 1.658,
+      "step": 13619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8047665357589722,
+      "learning_rate": 4.6402326196140636e-05,
+      "loss": 1.444,
+      "step": 13620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8347220420837402,
+      "learning_rate": 4.64020618556701e-05,
+      "loss": 1.4577,
+      "step": 13621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0155601501464844,
+      "learning_rate": 4.640179751519958e-05,
+      "loss": 1.4175,
+      "step": 13622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4887216091156006,
+      "learning_rate": 4.640153317472905e-05,
+      "loss": 1.3115,
+      "step": 13623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.048948287963867,
+      "learning_rate": 4.640126883425853e-05,
+      "loss": 1.6166,
+      "step": 13624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7411143779754639,
+      "learning_rate": 4.6401004493788003e-05,
+      "loss": 1.3659,
+      "step": 13625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2773849964141846,
+      "learning_rate": 4.640074015331747e-05,
+      "loss": 1.4854,
+      "step": 13626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1881821155548096,
+      "learning_rate": 4.640047581284695e-05,
+      "loss": 1.6176,
+      "step": 13627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3924351930618286,
+      "learning_rate": 4.640021147237642e-05,
+      "loss": 2.1846,
+      "step": 13628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7336339950561523,
+      "learning_rate": 4.63999471319059e-05,
+      "loss": 1.4555,
+      "step": 13629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5007905960083008,
+      "learning_rate": 4.639968279143537e-05,
+      "loss": 1.3154,
+      "step": 13630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.548671841621399,
+      "learning_rate": 4.6399418450964845e-05,
+      "loss": 1.3119,
+      "step": 13631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8656005859375,
+      "learning_rate": 4.639915411049432e-05,
+      "loss": 1.2451,
+      "step": 13632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8842809200286865,
+      "learning_rate": 4.639888977002379e-05,
+      "loss": 1.6046,
+      "step": 13633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5472623109817505,
+      "learning_rate": 4.6398625429553266e-05,
+      "loss": 1.2272,
+      "step": 13634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8305795192718506,
+      "learning_rate": 4.639836108908274e-05,
+      "loss": 1.8852,
+      "step": 13635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0398497581481934,
+      "learning_rate": 4.639809674861221e-05,
+      "loss": 1.4474,
+      "step": 13636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6976020336151123,
+      "learning_rate": 4.6397832408141686e-05,
+      "loss": 1.4715,
+      "step": 13637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.785619854927063,
+      "learning_rate": 4.639756806767116e-05,
+      "loss": 1.3918,
+      "step": 13638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8603286743164062,
+      "learning_rate": 4.6397303727200633e-05,
+      "loss": 1.2354,
+      "step": 13639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7621186971664429,
+      "learning_rate": 4.6397039386730114e-05,
+      "loss": 1.3348,
+      "step": 13640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0738065242767334,
+      "learning_rate": 4.639677504625958e-05,
+      "loss": 1.3926,
+      "step": 13641
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5739856958389282,
+      "learning_rate": 4.639651070578906e-05,
+      "loss": 1.2583,
+      "step": 13642
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6394524574279785,
+      "learning_rate": 4.639624636531853e-05,
+      "loss": 1.2266,
+      "step": 13643
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7662339210510254,
+      "learning_rate": 4.639598202484801e-05,
+      "loss": 1.4039,
+      "step": 13644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9531662464141846,
+      "learning_rate": 4.639571768437748e-05,
+      "loss": 1.4603,
+      "step": 13645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.664718747138977,
+      "learning_rate": 4.6395453343906955e-05,
+      "loss": 1.351,
+      "step": 13646
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1694507598876953,
+      "learning_rate": 4.639518900343643e-05,
+      "loss": 1.5441,
+      "step": 13647
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7176192998886108,
+      "learning_rate": 4.63949246629659e-05,
+      "loss": 1.274,
+      "step": 13648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9436501264572144,
+      "learning_rate": 4.6394660322495376e-05,
+      "loss": 1.4524,
+      "step": 13649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.936786413192749,
+      "learning_rate": 4.639439598202485e-05,
+      "loss": 1.4766,
+      "step": 13650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9442486763000488,
+      "learning_rate": 4.639413164155432e-05,
+      "loss": 1.4236,
+      "step": 13651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1448450088500977,
+      "learning_rate": 4.6393867301083797e-05,
+      "loss": 1.4834,
+      "step": 13652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7312861680984497,
+      "learning_rate": 4.639360296061327e-05,
+      "loss": 1.3805,
+      "step": 13653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7572931051254272,
+      "learning_rate": 4.6393338620142744e-05,
+      "loss": 2.2508,
+      "step": 13654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7950811386108398,
+      "learning_rate": 4.6393074279672224e-05,
+      "loss": 1.3,
+      "step": 13655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1274728775024414,
+      "learning_rate": 4.639280993920169e-05,
+      "loss": 1.3382,
+      "step": 13656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9558584690093994,
+      "learning_rate": 4.639254559873117e-05,
+      "loss": 1.3979,
+      "step": 13657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.648095965385437,
+      "learning_rate": 4.639228125826064e-05,
+      "loss": 1.3311,
+      "step": 13658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8554259538650513,
+      "learning_rate": 4.639201691779012e-05,
+      "loss": 1.5671,
+      "step": 13659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4705252647399902,
+      "learning_rate": 4.639175257731959e-05,
+      "loss": 1.2874,
+      "step": 13660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0268452167510986,
+      "learning_rate": 4.6391488236849066e-05,
+      "loss": 1.4654,
+      "step": 13661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4283127784729004,
+      "learning_rate": 4.639122389637854e-05,
+      "loss": 1.573,
+      "step": 13662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.837078332901001,
+      "learning_rate": 4.6390959555908006e-05,
+      "loss": 1.5141,
+      "step": 13663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9639456272125244,
+      "learning_rate": 4.6390695215437486e-05,
+      "loss": 1.5146,
+      "step": 13664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.609471082687378,
+      "learning_rate": 4.639043087496696e-05,
+      "loss": 1.2966,
+      "step": 13665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.837339997291565,
+      "learning_rate": 4.639016653449643e-05,
+      "loss": 1.4893,
+      "step": 13666
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6042770147323608,
+      "learning_rate": 4.638990219402591e-05,
+      "loss": 1.4445,
+      "step": 13667
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8246723413467407,
+      "learning_rate": 4.638963785355538e-05,
+      "loss": 1.3695,
+      "step": 13668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4964570999145508,
+      "learning_rate": 4.6389373513084854e-05,
+      "loss": 1.3569,
+      "step": 13669
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9847737550735474,
+      "learning_rate": 4.6389109172614334e-05,
+      "loss": 1.4125,
+      "step": 13670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.622257709503174,
+      "learning_rate": 4.63888448321438e-05,
+      "loss": 1.4194,
+      "step": 13671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2835371494293213,
+      "learning_rate": 4.6388580491673275e-05,
+      "loss": 1.1829,
+      "step": 13672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.395890235900879,
+      "learning_rate": 4.638831615120275e-05,
+      "loss": 1.3313,
+      "step": 13673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8708029985427856,
+      "learning_rate": 4.638805181073222e-05,
+      "loss": 1.4367,
+      "step": 13674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4888696670532227,
+      "learning_rate": 4.63877874702617e-05,
+      "loss": 1.37,
+      "step": 13675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6341214179992676,
+      "learning_rate": 4.638752312979117e-05,
+      "loss": 2.1194,
+      "step": 13676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5404309034347534,
+      "learning_rate": 4.638725878932065e-05,
+      "loss": 1.2536,
+      "step": 13677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078763961791992,
+      "learning_rate": 4.6386994448850116e-05,
+      "loss": 1.438,
+      "step": 13678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2550787925720215,
+      "learning_rate": 4.6386730108379597e-05,
+      "loss": 1.5257,
+      "step": 13679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8685603141784668,
+      "learning_rate": 4.638646576790907e-05,
+      "loss": 1.4189,
+      "step": 13680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9018337726593018,
+      "learning_rate": 4.6386201427438544e-05,
+      "loss": 1.3495,
+      "step": 13681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5444066524505615,
+      "learning_rate": 4.638593708696802e-05,
+      "loss": 1.2812,
+      "step": 13682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7900761365890503,
+      "learning_rate": 4.638567274649749e-05,
+      "loss": 1.3245,
+      "step": 13683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9641950130462646,
+      "learning_rate": 4.6385408406026964e-05,
+      "loss": 1.488,
+      "step": 13684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3504397869110107,
+      "learning_rate": 4.638514406555644e-05,
+      "loss": 1.4992,
+      "step": 13685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7089860439300537,
+      "learning_rate": 4.638487972508591e-05,
+      "loss": 1.4329,
+      "step": 13686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.031282901763916,
+      "learning_rate": 4.6384615384615385e-05,
+      "loss": 1.3567,
+      "step": 13687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.559377670288086,
+      "learning_rate": 4.638435104414486e-05,
+      "loss": 1.4967,
+      "step": 13688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5817184448242188,
+      "learning_rate": 4.638408670367433e-05,
+      "loss": 1.2936,
+      "step": 13689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1755242347717285,
+      "learning_rate": 4.638382236320381e-05,
+      "loss": 1.4367,
+      "step": 13690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.50821053981781,
+      "learning_rate": 4.638355802273328e-05,
+      "loss": 1.3653,
+      "step": 13691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7694191932678223,
+      "learning_rate": 4.638329368226276e-05,
+      "loss": 1.3998,
+      "step": 13692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7527278661727905,
+      "learning_rate": 4.6383029341792227e-05,
+      "loss": 1.414,
+      "step": 13693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.772032618522644,
+      "learning_rate": 4.638276500132171e-05,
+      "loss": 1.5314,
+      "step": 13694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.829890251159668,
+      "learning_rate": 4.638250066085118e-05,
+      "loss": 1.3692,
+      "step": 13695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8813589811325073,
+      "learning_rate": 4.6382236320380654e-05,
+      "loss": 1.3967,
+      "step": 13696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.72257661819458,
+      "learning_rate": 4.638197197991013e-05,
+      "loss": 1.4625,
+      "step": 13697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9258363246917725,
+      "learning_rate": 4.6381707639439594e-05,
+      "loss": 1.8446,
+      "step": 13698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6081174612045288,
+      "learning_rate": 4.6381443298969075e-05,
+      "loss": 1.3838,
+      "step": 13699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6704872846603394,
+      "learning_rate": 4.638117895849855e-05,
+      "loss": 1.4153,
+      "step": 13700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6645708084106445,
+      "learning_rate": 4.638091461802802e-05,
+      "loss": 1.3611,
+      "step": 13701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0761606693267822,
+      "learning_rate": 4.6380650277557495e-05,
+      "loss": 1.4574,
+      "step": 13702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7898337841033936,
+      "learning_rate": 4.638038593708697e-05,
+      "loss": 1.3952,
+      "step": 13703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9362949132919312,
+      "learning_rate": 4.638012159661644e-05,
+      "loss": 1.5366,
+      "step": 13704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7033308744430542,
+      "learning_rate": 4.637985725614592e-05,
+      "loss": 1.4941,
+      "step": 13705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0381925106048584,
+      "learning_rate": 4.637959291567539e-05,
+      "loss": 1.6189,
+      "step": 13706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9646199941635132,
+      "learning_rate": 4.637932857520487e-05,
+      "loss": 1.3102,
+      "step": 13707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.787198781967163,
+      "learning_rate": 4.637906423473434e-05,
+      "loss": 1.3038,
+      "step": 13708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7805817127227783,
+      "learning_rate": 4.637879989426381e-05,
+      "loss": 1.3782,
+      "step": 13709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4331716299057007,
+      "learning_rate": 4.637853555379329e-05,
+      "loss": 1.3041,
+      "step": 13710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5411189794540405,
+      "learning_rate": 4.637827121332276e-05,
+      "loss": 1.3367,
+      "step": 13711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8598477840423584,
+      "learning_rate": 4.637800687285224e-05,
+      "loss": 1.2232,
+      "step": 13712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9295791387557983,
+      "learning_rate": 4.6377742532381705e-05,
+      "loss": 1.552,
+      "step": 13713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7336704730987549,
+      "learning_rate": 4.6377478191911185e-05,
+      "loss": 1.2984,
+      "step": 13714
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.373456358909607,
+      "learning_rate": 4.637721385144066e-05,
+      "loss": 1.3016,
+      "step": 13715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.418299674987793,
+      "learning_rate": 4.637694951097013e-05,
+      "loss": 1.3277,
+      "step": 13716
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8127509355545044,
+      "learning_rate": 4.6376685170499606e-05,
+      "loss": 1.6092,
+      "step": 13717
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4993647336959839,
+      "learning_rate": 4.637642083002908e-05,
+      "loss": 1.24,
+      "step": 13718
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7319282293319702,
+      "learning_rate": 4.637615648955855e-05,
+      "loss": 1.3651,
+      "step": 13719
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.07895565032959,
+      "learning_rate": 4.6375892149088027e-05,
+      "loss": 1.4018,
+      "step": 13720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4048806428909302,
+      "learning_rate": 4.63756278086175e-05,
+      "loss": 1.3891,
+      "step": 13721
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5152069330215454,
+      "learning_rate": 4.6375363468146974e-05,
+      "loss": 1.4099,
+      "step": 13722
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.063171625137329,
+      "learning_rate": 4.637509912767645e-05,
+      "loss": 1.5309,
+      "step": 13723
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0892539024353027,
+      "learning_rate": 4.637483478720592e-05,
+      "loss": 1.4013,
+      "step": 13724
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0128843784332275,
+      "learning_rate": 4.63745704467354e-05,
+      "loss": 1.4057,
+      "step": 13725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9776619672775269,
+      "learning_rate": 4.637430610626487e-05,
+      "loss": 1.447,
+      "step": 13726
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7526116371154785,
+      "learning_rate": 4.637404176579435e-05,
+      "loss": 1.5208,
+      "step": 13727
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9408588409423828,
+      "learning_rate": 4.6373777425323815e-05,
+      "loss": 1.4837,
+      "step": 13728
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6905629634857178,
+      "learning_rate": 4.6373513084853295e-05,
+      "loss": 1.3775,
+      "step": 13729
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9519081115722656,
+      "learning_rate": 4.637324874438277e-05,
+      "loss": 1.6476,
+      "step": 13730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1166296005249023,
+      "learning_rate": 4.637298440391224e-05,
+      "loss": 1.3829,
+      "step": 13731
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8690556287765503,
+      "learning_rate": 4.6372720063441716e-05,
+      "loss": 1.1709,
+      "step": 13732
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2097220420837402,
+      "learning_rate": 4.637245572297119e-05,
+      "loss": 1.5642,
+      "step": 13733
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.516861915588379,
+      "learning_rate": 4.637219138250066e-05,
+      "loss": 1.2817,
+      "step": 13734
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.641404628753662,
+      "learning_rate": 4.637192704203014e-05,
+      "loss": 1.2504,
+      "step": 13735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9886200428009033,
+      "learning_rate": 4.637166270155961e-05,
+      "loss": 1.3519,
+      "step": 13736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6963180303573608,
+      "learning_rate": 4.6371398361089084e-05,
+      "loss": 1.3777,
+      "step": 13737
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7374690771102905,
+      "learning_rate": 4.637113402061856e-05,
+      "loss": 1.5278,
+      "step": 13738
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0603575706481934,
+      "learning_rate": 4.637086968014803e-05,
+      "loss": 1.4443,
+      "step": 13739
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.207304000854492,
+      "learning_rate": 4.637060533967751e-05,
+      "loss": 1.361,
+      "step": 13740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.819566011428833,
+      "learning_rate": 4.637034099920698e-05,
+      "loss": 1.4193,
+      "step": 13741
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9424102306365967,
+      "learning_rate": 4.637007665873646e-05,
+      "loss": 1.4987,
+      "step": 13742
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4906020164489746,
+      "learning_rate": 4.6369812318265925e-05,
+      "loss": 1.3666,
+      "step": 13743
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.769641637802124,
+      "learning_rate": 4.63695479777954e-05,
+      "loss": 1.3575,
+      "step": 13744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8890271186828613,
+      "learning_rate": 4.636928363732488e-05,
+      "loss": 1.3425,
+      "step": 13745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8819249868392944,
+      "learning_rate": 4.6369019296854346e-05,
+      "loss": 1.4334,
+      "step": 13746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1331496238708496,
+      "learning_rate": 4.6368754956383826e-05,
+      "loss": 1.5046,
+      "step": 13747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8716119527816772,
+      "learning_rate": 4.636849061591329e-05,
+      "loss": 1.3314,
+      "step": 13748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9525443315505981,
+      "learning_rate": 4.6368226275442774e-05,
+      "loss": 1.3924,
+      "step": 13749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7009373903274536,
+      "learning_rate": 4.636796193497225e-05,
+      "loss": 1.309,
+      "step": 13750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1138482093811035,
+      "learning_rate": 4.636769759450172e-05,
+      "loss": 1.6857,
+      "step": 13751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9610353708267212,
+      "learning_rate": 4.6367433254031194e-05,
+      "loss": 1.5162,
+      "step": 13752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4853649139404297,
+      "learning_rate": 4.636716891356067e-05,
+      "loss": 1.2466,
+      "step": 13753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4985541105270386,
+      "learning_rate": 4.636690457309014e-05,
+      "loss": 1.3898,
+      "step": 13754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8942182064056396,
+      "learning_rate": 4.6366640232619615e-05,
+      "loss": 1.4599,
+      "step": 13755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5643433332443237,
+      "learning_rate": 4.636637589214909e-05,
+      "loss": 1.2803,
+      "step": 13756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.479646921157837,
+      "learning_rate": 4.636611155167856e-05,
+      "loss": 1.7259,
+      "step": 13757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.244481325149536,
+      "learning_rate": 4.6365847211208036e-05,
+      "loss": 1.4942,
+      "step": 13758
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5546715259552002,
+      "learning_rate": 4.636558287073751e-05,
+      "loss": 1.5614,
+      "step": 13759
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.615185499191284,
+      "learning_rate": 4.636531853026699e-05,
+      "loss": 1.7216,
+      "step": 13760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8698033094406128,
+      "learning_rate": 4.6365054189796456e-05,
+      "loss": 1.3146,
+      "step": 13761
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9550055265426636,
+      "learning_rate": 4.636478984932594e-05,
+      "loss": 1.458,
+      "step": 13762
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7164368629455566,
+      "learning_rate": 4.6364525508855404e-05,
+      "loss": 1.4099,
+      "step": 13763
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.567426323890686,
+      "learning_rate": 4.6364261168384884e-05,
+      "loss": 1.2344,
+      "step": 13764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.126680374145508,
+      "learning_rate": 4.636399682791436e-05,
+      "loss": 1.4394,
+      "step": 13765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8058626651763916,
+      "learning_rate": 4.636373248744383e-05,
+      "loss": 1.3149,
+      "step": 13766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.983636498451233,
+      "learning_rate": 4.6363468146973305e-05,
+      "loss": 1.4472,
+      "step": 13767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.318453788757324,
+      "learning_rate": 4.636320380650278e-05,
+      "loss": 1.4858,
+      "step": 13768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5214345455169678,
+      "learning_rate": 4.636293946603225e-05,
+      "loss": 1.4077,
+      "step": 13769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7108852863311768,
+      "learning_rate": 4.6362675125561725e-05,
+      "loss": 1.4105,
+      "step": 13770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9351252317428589,
+      "learning_rate": 4.63624107850912e-05,
+      "loss": 1.4647,
+      "step": 13771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9067580699920654,
+      "learning_rate": 4.636214644462067e-05,
+      "loss": 1.3673,
+      "step": 13772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6984535455703735,
+      "learning_rate": 4.6361882104150146e-05,
+      "loss": 1.3727,
+      "step": 13773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7565314769744873,
+      "learning_rate": 4.636161776367962e-05,
+      "loss": 1.2547,
+      "step": 13774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4282394647598267,
+      "learning_rate": 4.63613534232091e-05,
+      "loss": 1.1854,
+      "step": 13775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5673080682754517,
+      "learning_rate": 4.636108908273857e-05,
+      "loss": 1.3957,
+      "step": 13776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8797751665115356,
+      "learning_rate": 4.636082474226805e-05,
+      "loss": 1.513,
+      "step": 13777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.826532006263733,
+      "learning_rate": 4.6360560401797514e-05,
+      "loss": 1.3833,
+      "step": 13778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5579966306686401,
+      "learning_rate": 4.6360296061326994e-05,
+      "loss": 1.2747,
+      "step": 13779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7603579759597778,
+      "learning_rate": 4.636003172085647e-05,
+      "loss": 1.3771,
+      "step": 13780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5391780138015747,
+      "learning_rate": 4.6359767380385935e-05,
+      "loss": 1.4072,
+      "step": 13781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6622815132141113,
+      "learning_rate": 4.6359503039915415e-05,
+      "loss": 1.482,
+      "step": 13782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6235976219177246,
+      "learning_rate": 4.635923869944488e-05,
+      "loss": 1.4486,
+      "step": 13783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4653466939926147,
+      "learning_rate": 4.635897435897436e-05,
+      "loss": 1.2584,
+      "step": 13784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.100250005722046,
+      "learning_rate": 4.6358710018503836e-05,
+      "loss": 1.5218,
+      "step": 13785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7078216075897217,
+      "learning_rate": 4.635844567803331e-05,
+      "loss": 1.3601,
+      "step": 13786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5486456155776978,
+      "learning_rate": 4.635818133756278e-05,
+      "loss": 1.2987,
+      "step": 13787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0283920764923096,
+      "learning_rate": 4.6357916997092256e-05,
+      "loss": 1.4466,
+      "step": 13788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.877089023590088,
+      "learning_rate": 4.635765265662173e-05,
+      "loss": 1.4033,
+      "step": 13789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7085851430892944,
+      "learning_rate": 4.6357388316151204e-05,
+      "loss": 1.8815,
+      "step": 13790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4443440437316895,
+      "learning_rate": 4.635712397568068e-05,
+      "loss": 1.2909,
+      "step": 13791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.294581413269043,
+      "learning_rate": 4.635685963521015e-05,
+      "loss": 1.6315,
+      "step": 13792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4849460124969482,
+      "learning_rate": 4.6356595294739624e-05,
+      "loss": 1.4992,
+      "step": 13793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5219098329544067,
+      "learning_rate": 4.63563309542691e-05,
+      "loss": 1.3577,
+      "step": 13794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9066420793533325,
+      "learning_rate": 4.635606661379858e-05,
+      "loss": 1.4371,
+      "step": 13795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7478663921356201,
+      "learning_rate": 4.6355802273328045e-05,
+      "loss": 1.4289,
+      "step": 13796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.825259804725647,
+      "learning_rate": 4.6355537932857525e-05,
+      "loss": 1.5954,
+      "step": 13797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6494501829147339,
+      "learning_rate": 4.635527359238699e-05,
+      "loss": 1.4905,
+      "step": 13798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0276670455932617,
+      "learning_rate": 4.635500925191647e-05,
+      "loss": 1.5491,
+      "step": 13799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2561888694763184,
+      "learning_rate": 4.6354744911445946e-05,
+      "loss": 1.5015,
+      "step": 13800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7188866138458252,
+      "learning_rate": 4.635448057097542e-05,
+      "loss": 1.421,
+      "step": 13801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6627776622772217,
+      "learning_rate": 4.635421623050489e-05,
+      "loss": 1.306,
+      "step": 13802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9019463062286377,
+      "learning_rate": 4.635395189003437e-05,
+      "loss": 1.4734,
+      "step": 13803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6949961185455322,
+      "learning_rate": 4.635368754956384e-05,
+      "loss": 1.3164,
+      "step": 13804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8931283950805664,
+      "learning_rate": 4.6353423209093314e-05,
+      "loss": 1.3699,
+      "step": 13805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8431129455566406,
+      "learning_rate": 4.635315886862279e-05,
+      "loss": 1.4553,
+      "step": 13806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1530232429504395,
+      "learning_rate": 4.635289452815226e-05,
+      "loss": 1.6853,
+      "step": 13807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.722716212272644,
+      "learning_rate": 4.6352630187681735e-05,
+      "loss": 1.4064,
+      "step": 13808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7480944395065308,
+      "learning_rate": 4.635236584721121e-05,
+      "loss": 1.2762,
+      "step": 13809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5502649545669556,
+      "learning_rate": 4.635210150674069e-05,
+      "loss": 1.4219,
+      "step": 13810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6060415506362915,
+      "learning_rate": 4.6351837166270155e-05,
+      "loss": 1.3475,
+      "step": 13811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8612006902694702,
+      "learning_rate": 4.6351572825799636e-05,
+      "loss": 1.438,
+      "step": 13812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8738480806350708,
+      "learning_rate": 4.63513084853291e-05,
+      "loss": 1.5247,
+      "step": 13813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6623022556304932,
+      "learning_rate": 4.635104414485858e-05,
+      "loss": 1.3546,
+      "step": 13814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.778733491897583,
+      "learning_rate": 4.6350779804388056e-05,
+      "loss": 1.4044,
+      "step": 13815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6245301961898804,
+      "learning_rate": 4.635051546391752e-05,
+      "loss": 1.381,
+      "step": 13816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.547013998031616,
+      "learning_rate": 4.6350251123447004e-05,
+      "loss": 1.4747,
+      "step": 13817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3921525478363037,
+      "learning_rate": 4.634998678297647e-05,
+      "loss": 2.615,
+      "step": 13818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4832336902618408,
+      "learning_rate": 4.634972244250595e-05,
+      "loss": 1.9725,
+      "step": 13819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5812249183654785,
+      "learning_rate": 4.6349458102035424e-05,
+      "loss": 1.5262,
+      "step": 13820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.647466778755188,
+      "learning_rate": 4.63491937615649e-05,
+      "loss": 1.2007,
+      "step": 13821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6569957733154297,
+      "learning_rate": 4.634892942109437e-05,
+      "loss": 1.457,
+      "step": 13822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.508223533630371,
+      "learning_rate": 4.6348665080623845e-05,
+      "loss": 1.3073,
+      "step": 13823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3566538095474243,
+      "learning_rate": 4.634840074015332e-05,
+      "loss": 1.339,
+      "step": 13824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0171780586242676,
+      "learning_rate": 4.63481363996828e-05,
+      "loss": 2.4474,
+      "step": 13825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.032100200653076,
+      "learning_rate": 4.6347872059212266e-05,
+      "loss": 1.3465,
+      "step": 13826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8637828826904297,
+      "learning_rate": 4.634760771874174e-05,
+      "loss": 1.3581,
+      "step": 13827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6443065404891968,
+      "learning_rate": 4.634734337827121e-05,
+      "loss": 1.2815,
+      "step": 13828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4651107788085938,
+      "learning_rate": 4.6347079037800686e-05,
+      "loss": 1.2842,
+      "step": 13829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8778218030929565,
+      "learning_rate": 4.634681469733017e-05,
+      "loss": 1.9722,
+      "step": 13830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2629811763763428,
+      "learning_rate": 4.6346550356859634e-05,
+      "loss": 1.3787,
+      "step": 13831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7660294771194458,
+      "learning_rate": 4.6346286016389114e-05,
+      "loss": 1.295,
+      "step": 13832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3565531969070435,
+      "learning_rate": 4.634602167591858e-05,
+      "loss": 2.1755,
+      "step": 13833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7753714323043823,
+      "learning_rate": 4.634575733544806e-05,
+      "loss": 2.3738,
+      "step": 13834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5624094009399414,
+      "learning_rate": 4.6345492994977535e-05,
+      "loss": 1.1622,
+      "step": 13835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.517207145690918,
+      "learning_rate": 4.634522865450701e-05,
+      "loss": 1.3461,
+      "step": 13836
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.202115058898926,
+      "learning_rate": 4.634496431403648e-05,
+      "loss": 1.5826,
+      "step": 13837
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7410086393356323,
+      "learning_rate": 4.6344699973565955e-05,
+      "loss": 1.4274,
+      "step": 13838
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.515405535697937,
+      "learning_rate": 4.634443563309543e-05,
+      "loss": 1.3303,
+      "step": 13839
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9508819580078125,
+      "learning_rate": 4.63441712926249e-05,
+      "loss": 1.4776,
+      "step": 13840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.755462646484375,
+      "learning_rate": 4.6343906952154376e-05,
+      "loss": 1.5062,
+      "step": 13841
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.706879734992981,
+      "learning_rate": 4.634364261168385e-05,
+      "loss": 1.3641,
+      "step": 13842
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.377904176712036,
+      "learning_rate": 4.634337827121332e-05,
+      "loss": 1.3723,
+      "step": 13843
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6877450942993164,
+      "learning_rate": 4.63431139307428e-05,
+      "loss": 1.4197,
+      "step": 13844
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9360291957855225,
+      "learning_rate": 4.634284959027228e-05,
+      "loss": 1.4982,
+      "step": 13845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.5334625244140625,
+      "learning_rate": 4.6342585249801744e-05,
+      "loss": 1.6691,
+      "step": 13846
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.505455255508423,
+      "learning_rate": 4.6342320909331224e-05,
+      "loss": 1.5028,
+      "step": 13847
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4635249376296997,
+      "learning_rate": 4.634205656886069e-05,
+      "loss": 2.2889,
+      "step": 13848
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.718220591545105,
+      "learning_rate": 4.634179222839017e-05,
+      "loss": 1.3257,
+      "step": 13849
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3156726360321045,
+      "learning_rate": 4.6341527887919645e-05,
+      "loss": 1.5202,
+      "step": 13850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.244823932647705,
+      "learning_rate": 4.634126354744912e-05,
+      "loss": 1.5236,
+      "step": 13851
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9068045616149902,
+      "learning_rate": 4.634099920697859e-05,
+      "loss": 1.7087,
+      "step": 13852
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2287585735321045,
+      "learning_rate": 4.634073486650806e-05,
+      "loss": 1.4177,
+      "step": 13853
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5622023344039917,
+      "learning_rate": 4.634047052603754e-05,
+      "loss": 1.2754,
+      "step": 13854
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6080429553985596,
+      "learning_rate": 4.634020618556701e-05,
+      "loss": 1.2345,
+      "step": 13855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6601649522781372,
+      "learning_rate": 4.6339941845096486e-05,
+      "loss": 1.3745,
+      "step": 13856
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7038378715515137,
+      "learning_rate": 4.633967750462596e-05,
+      "loss": 1.3762,
+      "step": 13857
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.273329019546509,
+      "learning_rate": 4.6339413164155434e-05,
+      "loss": 1.3943,
+      "step": 13858
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.249354124069214,
+      "learning_rate": 4.633914882368491e-05,
+      "loss": 1.4785,
+      "step": 13859
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5678465366363525,
+      "learning_rate": 4.633888448321439e-05,
+      "loss": 1.3622,
+      "step": 13860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.002368450164795,
+      "learning_rate": 4.6338620142743854e-05,
+      "loss": 1.4773,
+      "step": 13861
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6468980312347412,
+      "learning_rate": 4.633835580227333e-05,
+      "loss": 1.4318,
+      "step": 13862
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2209055423736572,
+      "learning_rate": 4.63380914618028e-05,
+      "loss": 1.3871,
+      "step": 13863
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7223228216171265,
+      "learning_rate": 4.6337827121332275e-05,
+      "loss": 1.4423,
+      "step": 13864
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6750805377960205,
+      "learning_rate": 4.6337562780861755e-05,
+      "loss": 1.3453,
+      "step": 13865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1805813312530518,
+      "learning_rate": 4.633729844039122e-05,
+      "loss": 1.5518,
+      "step": 13866
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.732238531112671,
+      "learning_rate": 4.63370340999207e-05,
+      "loss": 1.3456,
+      "step": 13867
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.315180778503418,
+      "learning_rate": 4.633676975945017e-05,
+      "loss": 1.4595,
+      "step": 13868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5861265659332275,
+      "learning_rate": 4.633650541897965e-05,
+      "loss": 1.3497,
+      "step": 13869
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1742827892303467,
+      "learning_rate": 4.633624107850912e-05,
+      "loss": 1.5152,
+      "step": 13870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.659339189529419,
+      "learning_rate": 4.63359767380386e-05,
+      "loss": 1.4037,
+      "step": 13871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5172582864761353,
+      "learning_rate": 4.633571239756807e-05,
+      "loss": 1.3206,
+      "step": 13872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6389237642288208,
+      "learning_rate": 4.6335448057097544e-05,
+      "loss": 1.3837,
+      "step": 13873
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.139486312866211,
+      "learning_rate": 4.633518371662702e-05,
+      "loss": 1.569,
+      "step": 13874
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7102890014648438,
+      "learning_rate": 4.633491937615649e-05,
+      "loss": 1.3102,
+      "step": 13875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5060513019561768,
+      "learning_rate": 4.6334655035685965e-05,
+      "loss": 1.454,
+      "step": 13876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5169357061386108,
+      "learning_rate": 4.633439069521544e-05,
+      "loss": 1.2843,
+      "step": 13877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.932072162628174,
+      "learning_rate": 4.633412635474491e-05,
+      "loss": 1.4434,
+      "step": 13878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2831521034240723,
+      "learning_rate": 4.6333862014274385e-05,
+      "loss": 2.5402,
+      "step": 13879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5346038341522217,
+      "learning_rate": 4.6333597673803866e-05,
+      "loss": 1.4352,
+      "step": 13880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.237797975540161,
+      "learning_rate": 4.633333333333333e-05,
+      "loss": 1.4404,
+      "step": 13881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.304006576538086,
+      "learning_rate": 4.633306899286281e-05,
+      "loss": 1.5099,
+      "step": 13882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9761531352996826,
+      "learning_rate": 4.633280465239228e-05,
+      "loss": 1.4422,
+      "step": 13883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4383621215820312,
+      "learning_rate": 4.633254031192176e-05,
+      "loss": 1.3318,
+      "step": 13884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.452894926071167,
+      "learning_rate": 4.6332275971451233e-05,
+      "loss": 1.4019,
+      "step": 13885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8782868385314941,
+      "learning_rate": 4.633201163098071e-05,
+      "loss": 1.4839,
+      "step": 13886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3520362377166748,
+      "learning_rate": 4.633174729051018e-05,
+      "loss": 1.2079,
+      "step": 13887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6735421419143677,
+      "learning_rate": 4.633148295003965e-05,
+      "loss": 1.2873,
+      "step": 13888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6798808574676514,
+      "learning_rate": 4.633121860956913e-05,
+      "loss": 1.3564,
+      "step": 13889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.241711378097534,
+      "learning_rate": 4.63309542690986e-05,
+      "loss": 1.5225,
+      "step": 13890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5498708486557007,
+      "learning_rate": 4.6330689928628075e-05,
+      "loss": 1.2861,
+      "step": 13891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.565040946006775,
+      "learning_rate": 4.633042558815755e-05,
+      "loss": 1.3734,
+      "step": 13892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5727423429489136,
+      "learning_rate": 4.633016124768702e-05,
+      "loss": 1.7652,
+      "step": 13893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4947298765182495,
+      "learning_rate": 4.6329896907216496e-05,
+      "loss": 1.3301,
+      "step": 13894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6894242763519287,
+      "learning_rate": 4.6329632566745976e-05,
+      "loss": 1.4193,
+      "step": 13895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4837197065353394,
+      "learning_rate": 4.632936822627544e-05,
+      "loss": 2.3184,
+      "step": 13896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0016403198242188,
+      "learning_rate": 4.632910388580492e-05,
+      "loss": 1.375,
+      "step": 13897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.873162031173706,
+      "learning_rate": 4.632883954533439e-05,
+      "loss": 1.554,
+      "step": 13898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5731287002563477,
+      "learning_rate": 4.6328575204863863e-05,
+      "loss": 1.3151,
+      "step": 13899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.176016330718994,
+      "learning_rate": 4.6328310864393344e-05,
+      "loss": 1.5232,
+      "step": 13900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3744866847991943,
+      "learning_rate": 4.632804652392281e-05,
+      "loss": 1.6395,
+      "step": 13901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5897799730300903,
+      "learning_rate": 4.632778218345229e-05,
+      "loss": 2.0328,
+      "step": 13902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5946882963180542,
+      "learning_rate": 4.632751784298176e-05,
+      "loss": 1.4681,
+      "step": 13903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8725310564041138,
+      "learning_rate": 4.632725350251124e-05,
+      "loss": 1.64,
+      "step": 13904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6183664798736572,
+      "learning_rate": 4.632698916204071e-05,
+      "loss": 1.5431,
+      "step": 13905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8586416244506836,
+      "learning_rate": 4.6326724821570185e-05,
+      "loss": 1.4256,
+      "step": 13906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7579536437988281,
+      "learning_rate": 4.632646048109966e-05,
+      "loss": 1.4073,
+      "step": 13907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.594337821006775,
+      "learning_rate": 4.632619614062913e-05,
+      "loss": 1.3595,
+      "step": 13908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5928370952606201,
+      "learning_rate": 4.6325931800158606e-05,
+      "loss": 1.3665,
+      "step": 13909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8645318746566772,
+      "learning_rate": 4.632566745968808e-05,
+      "loss": 1.5196,
+      "step": 13910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8825550079345703,
+      "learning_rate": 4.632540311921755e-05,
+      "loss": 1.3042,
+      "step": 13911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8848069906234741,
+      "learning_rate": 4.632513877874703e-05,
+      "loss": 1.5127,
+      "step": 13912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0065159797668457,
+      "learning_rate": 4.63248744382765e-05,
+      "loss": 1.4534,
+      "step": 13913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8425955772399902,
+      "learning_rate": 4.6324610097805974e-05,
+      "loss": 1.3628,
+      "step": 13914
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.18892765045166,
+      "learning_rate": 4.6324345757335454e-05,
+      "loss": 1.453,
+      "step": 13915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0622146129608154,
+      "learning_rate": 4.632408141686492e-05,
+      "loss": 1.2534,
+      "step": 13916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8783072233200073,
+      "learning_rate": 4.63238170763944e-05,
+      "loss": 1.4115,
+      "step": 13917
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.469435453414917,
+      "learning_rate": 4.632355273592387e-05,
+      "loss": 1.7452,
+      "step": 13918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5606918334960938,
+      "learning_rate": 4.632328839545335e-05,
+      "loss": 1.4595,
+      "step": 13919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.436231851577759,
+      "learning_rate": 4.632302405498282e-05,
+      "loss": 1.4765,
+      "step": 13920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.70108163356781,
+      "learning_rate": 4.6322759714512296e-05,
+      "loss": 1.4144,
+      "step": 13921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8602780103683472,
+      "learning_rate": 4.632249537404177e-05,
+      "loss": 1.4346,
+      "step": 13922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.937273621559143,
+      "learning_rate": 4.632223103357124e-05,
+      "loss": 1.5598,
+      "step": 13923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5414153337478638,
+      "learning_rate": 4.6321966693100716e-05,
+      "loss": 1.2516,
+      "step": 13924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8143043518066406,
+      "learning_rate": 4.632170235263019e-05,
+      "loss": 1.4148,
+      "step": 13925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5911834239959717,
+      "learning_rate": 4.6321438012159663e-05,
+      "loss": 1.3051,
+      "step": 13926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28362774848938,
+      "learning_rate": 4.632117367168914e-05,
+      "loss": 1.4481,
+      "step": 13927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.445595383644104,
+      "learning_rate": 4.632090933121861e-05,
+      "loss": 1.244,
+      "step": 13928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7990058660507202,
+      "learning_rate": 4.6320644990748084e-05,
+      "loss": 1.361,
+      "step": 13929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4478299617767334,
+      "learning_rate": 4.6320380650277565e-05,
+      "loss": 1.4129,
+      "step": 13930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2513232231140137,
+      "learning_rate": 4.632011630980703e-05,
+      "loss": 1.6261,
+      "step": 13931
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6029677391052246,
+      "learning_rate": 4.631985196933651e-05,
+      "loss": 1.7444,
+      "step": 13932
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3951785564422607,
+      "learning_rate": 4.631958762886598e-05,
+      "loss": 1.5179,
+      "step": 13933
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0291292667388916,
+      "learning_rate": 4.631932328839545e-05,
+      "loss": 1.4441,
+      "step": 13934
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1190409660339355,
+      "learning_rate": 4.631905894792493e-05,
+      "loss": 1.4907,
+      "step": 13935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6447975635528564,
+      "learning_rate": 4.63187946074544e-05,
+      "loss": 1.3645,
+      "step": 13936
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.271851062774658,
+      "learning_rate": 4.631853026698388e-05,
+      "loss": 1.4474,
+      "step": 13937
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.398176431655884,
+      "learning_rate": 4.6318265926513346e-05,
+      "loss": 1.5452,
+      "step": 13938
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6237566471099854,
+      "learning_rate": 4.631800158604283e-05,
+      "loss": 1.5004,
+      "step": 13939
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.634685754776001,
+      "learning_rate": 4.63177372455723e-05,
+      "loss": 1.2507,
+      "step": 13940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5998677015304565,
+      "learning_rate": 4.6317472905101774e-05,
+      "loss": 1.55,
+      "step": 13941
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8098443746566772,
+      "learning_rate": 4.631720856463125e-05,
+      "loss": 1.3474,
+      "step": 13942
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9808893203735352,
+      "learning_rate": 4.631694422416072e-05,
+      "loss": 1.4097,
+      "step": 13943
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9383541345596313,
+      "learning_rate": 4.6316679883690195e-05,
+      "loss": 1.4033,
+      "step": 13944
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8394070863723755,
+      "learning_rate": 4.631641554321967e-05,
+      "loss": 1.4362,
+      "step": 13945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.713597297668457,
+      "learning_rate": 4.631615120274914e-05,
+      "loss": 1.2867,
+      "step": 13946
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171616315841675,
+      "learning_rate": 4.6315886862278615e-05,
+      "loss": 1.3387,
+      "step": 13947
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9281911849975586,
+      "learning_rate": 4.631562252180809e-05,
+      "loss": 1.561,
+      "step": 13948
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.04483699798584,
+      "learning_rate": 4.631535818133756e-05,
+      "loss": 1.5477,
+      "step": 13949
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4208707809448242,
+      "learning_rate": 4.631509384086704e-05,
+      "loss": 1.3083,
+      "step": 13950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7378228902816772,
+      "learning_rate": 4.631482950039651e-05,
+      "loss": 1.241,
+      "step": 13951
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.781732439994812,
+      "learning_rate": 4.631456515992599e-05,
+      "loss": 1.3261,
+      "step": 13952
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8072144985198975,
+      "learning_rate": 4.631430081945546e-05,
+      "loss": 2.218,
+      "step": 13953
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0457592010498047,
+      "learning_rate": 4.631403647898494e-05,
+      "loss": 1.372,
+      "step": 13954
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8412691354751587,
+      "learning_rate": 4.631377213851441e-05,
+      "loss": 1.4716,
+      "step": 13955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5662546157836914,
+      "learning_rate": 4.6313507798043884e-05,
+      "loss": 1.3923,
+      "step": 13956
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.862072229385376,
+      "learning_rate": 4.631324345757336e-05,
+      "loss": 1.343,
+      "step": 13957
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6153875589370728,
+      "learning_rate": 4.631297911710283e-05,
+      "loss": 1.522,
+      "step": 13958
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.780754566192627,
+      "learning_rate": 4.6312714776632305e-05,
+      "loss": 1.3495,
+      "step": 13959
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8180302381515503,
+      "learning_rate": 4.631245043616178e-05,
+      "loss": 1.587,
+      "step": 13960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6900134086608887,
+      "learning_rate": 4.631218609569125e-05,
+      "loss": 1.4051,
+      "step": 13961
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3241504430770874,
+      "learning_rate": 4.6311921755220726e-05,
+      "loss": 1.3529,
+      "step": 13962
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5891733169555664,
+      "learning_rate": 4.63116574147502e-05,
+      "loss": 1.5313,
+      "step": 13963
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4074727296829224,
+      "learning_rate": 4.631139307427967e-05,
+      "loss": 1.4303,
+      "step": 13964
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8284274339675903,
+      "learning_rate": 4.631112873380915e-05,
+      "loss": 2.3093,
+      "step": 13965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6766119003295898,
+      "learning_rate": 4.631086439333862e-05,
+      "loss": 1.3726,
+      "step": 13966
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4887406826019287,
+      "learning_rate": 4.63106000528681e-05,
+      "loss": 1.308,
+      "step": 13967
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7259851694107056,
+      "learning_rate": 4.631033571239757e-05,
+      "loss": 2.5965,
+      "step": 13968
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.753312587738037,
+      "learning_rate": 4.631007137192705e-05,
+      "loss": 1.575,
+      "step": 13969
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8081705570220947,
+      "learning_rate": 4.630980703145652e-05,
+      "loss": 1.4329,
+      "step": 13970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0920629501342773,
+      "learning_rate": 4.630954269098599e-05,
+      "loss": 1.8419,
+      "step": 13971
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6121456623077393,
+      "learning_rate": 4.630927835051547e-05,
+      "loss": 1.3618,
+      "step": 13972
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7920081615447998,
+      "learning_rate": 4.6309014010044935e-05,
+      "loss": 1.2725,
+      "step": 13973
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6149110794067383,
+      "learning_rate": 4.6308749669574415e-05,
+      "loss": 1.4162,
+      "step": 13974
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.562793254852295,
+      "learning_rate": 4.630848532910389e-05,
+      "loss": 1.6566,
+      "step": 13975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.746160626411438,
+      "learning_rate": 4.630822098863336e-05,
+      "loss": 1.4007,
+      "step": 13976
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9798176288604736,
+      "learning_rate": 4.6307956648162836e-05,
+      "loss": 1.6553,
+      "step": 13977
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8675239086151123,
+      "learning_rate": 4.630769230769231e-05,
+      "loss": 1.455,
+      "step": 13978
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6219966411590576,
+      "learning_rate": 4.630742796722178e-05,
+      "loss": 1.3354,
+      "step": 13979
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7434186935424805,
+      "learning_rate": 4.630716362675126e-05,
+      "loss": 1.3779,
+      "step": 13980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.492767095565796,
+      "learning_rate": 4.630689928628073e-05,
+      "loss": 1.23,
+      "step": 13981
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8035856485366821,
+      "learning_rate": 4.6306634945810204e-05,
+      "loss": 1.3571,
+      "step": 13982
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.878713607788086,
+      "learning_rate": 4.630637060533968e-05,
+      "loss": 1.3643,
+      "step": 13983
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.042635202407837,
+      "learning_rate": 4.630610626486915e-05,
+      "loss": 1.3726,
+      "step": 13984
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6488136053085327,
+      "learning_rate": 4.630584192439863e-05,
+      "loss": 1.3425,
+      "step": 13985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1850314140319824,
+      "learning_rate": 4.63055775839281e-05,
+      "loss": 1.4411,
+      "step": 13986
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9548463821411133,
+      "learning_rate": 4.630531324345758e-05,
+      "loss": 1.3961,
+      "step": 13987
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.150942087173462,
+      "learning_rate": 4.6305048902987045e-05,
+      "loss": 1.3803,
+      "step": 13988
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8298462629318237,
+      "learning_rate": 4.6304784562516526e-05,
+      "loss": 1.3053,
+      "step": 13989
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5694440603256226,
+      "learning_rate": 4.6304520222046e-05,
+      "loss": 1.497,
+      "step": 13990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4784629344940186,
+      "learning_rate": 4.630425588157547e-05,
+      "loss": 1.3824,
+      "step": 13991
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.820357084274292,
+      "learning_rate": 4.6303991541104946e-05,
+      "loss": 1.4553,
+      "step": 13992
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9539437294006348,
+      "learning_rate": 4.630372720063442e-05,
+      "loss": 1.3812,
+      "step": 13993
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7857186794281006,
+      "learning_rate": 4.630346286016389e-05,
+      "loss": 1.4561,
+      "step": 13994
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7327117919921875,
+      "learning_rate": 4.630319851969337e-05,
+      "loss": 1.3846,
+      "step": 13995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5066241025924683,
+      "learning_rate": 4.630293417922284e-05,
+      "loss": 1.6435,
+      "step": 13996
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8995280265808105,
+      "learning_rate": 4.6302669838752314e-05,
+      "loss": 1.4079,
+      "step": 13997
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0602450370788574,
+      "learning_rate": 4.630240549828179e-05,
+      "loss": 1.4502,
+      "step": 13998
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6818890571594238,
+      "learning_rate": 4.630214115781126e-05,
+      "loss": 1.1407,
+      "step": 13999
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5587283372879028,
+      "learning_rate": 4.630187681734074e-05,
+      "loss": 1.3555,
+      "step": 14000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.641725778579712,
+      "learning_rate": 4.630161247687021e-05,
+      "loss": 1.2217,
+      "step": 14001
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.533684015274048,
+      "learning_rate": 4.630134813639969e-05,
+      "loss": 1.6075,
+      "step": 14002
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8808341026306152,
+      "learning_rate": 4.6301083795929156e-05,
+      "loss": 1.5698,
+      "step": 14003
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8774062395095825,
+      "learning_rate": 4.6300819455458636e-05,
+      "loss": 1.3702,
+      "step": 14004
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0256452560424805,
+      "learning_rate": 4.630055511498811e-05,
+      "loss": 1.4574,
+      "step": 14005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.063856601715088,
+      "learning_rate": 4.6300290774517576e-05,
+      "loss": 1.5066,
+      "step": 14006
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1513149738311768,
+      "learning_rate": 4.6300026434047057e-05,
+      "loss": 1.4804,
+      "step": 14007
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4509546756744385,
+      "learning_rate": 4.629976209357652e-05,
+      "loss": 1.5564,
+      "step": 14008
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6531082391738892,
+      "learning_rate": 4.6299497753106004e-05,
+      "loss": 1.3545,
+      "step": 14009
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3180699348449707,
+      "learning_rate": 4.629923341263548e-05,
+      "loss": 1.6144,
+      "step": 14010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.444286823272705,
+      "learning_rate": 4.629896907216495e-05,
+      "loss": 1.3857,
+      "step": 14011
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0939009189605713,
+      "learning_rate": 4.6298704731694424e-05,
+      "loss": 1.3526,
+      "step": 14012
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0811402797698975,
+      "learning_rate": 4.62984403912239e-05,
+      "loss": 1.7527,
+      "step": 14013
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5700246095657349,
+      "learning_rate": 4.629817605075337e-05,
+      "loss": 1.3946,
+      "step": 14014
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8810527324676514,
+      "learning_rate": 4.629791171028285e-05,
+      "loss": 1.4631,
+      "step": 14015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5754674673080444,
+      "learning_rate": 4.629764736981232e-05,
+      "loss": 1.2424,
+      "step": 14016
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.580107569694519,
+      "learning_rate": 4.629738302934179e-05,
+      "loss": 1.2947,
+      "step": 14017
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6512531042099,
+      "learning_rate": 4.6297118688871266e-05,
+      "loss": 1.4067,
+      "step": 14018
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.116398811340332,
+      "learning_rate": 4.629685434840074e-05,
+      "loss": 1.547,
+      "step": 14019
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.255117654800415,
+      "learning_rate": 4.629659000793022e-05,
+      "loss": 1.9722,
+      "step": 14020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8471920490264893,
+      "learning_rate": 4.6296325667459687e-05,
+      "loss": 1.4192,
+      "step": 14021
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.592176675796509,
+      "learning_rate": 4.629606132698917e-05,
+      "loss": 1.4355,
+      "step": 14022
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.117497682571411,
+      "learning_rate": 4.6295796986518634e-05,
+      "loss": 1.4198,
+      "step": 14023
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2453691959381104,
+      "learning_rate": 4.6295532646048114e-05,
+      "loss": 1.501,
+      "step": 14024
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0203166007995605,
+      "learning_rate": 4.629526830557759e-05,
+      "loss": 1.501,
+      "step": 14025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7765711545944214,
+      "learning_rate": 4.629500396510706e-05,
+      "loss": 1.3102,
+      "step": 14026
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6962875127792358,
+      "learning_rate": 4.6294739624636535e-05,
+      "loss": 1.2662,
+      "step": 14027
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.63718581199646,
+      "learning_rate": 4.629447528416601e-05,
+      "loss": 1.4701,
+      "step": 14028
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.98026704788208,
+      "learning_rate": 4.629421094369548e-05,
+      "loss": 1.361,
+      "step": 14029
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6429563760757446,
+      "learning_rate": 4.6293946603224955e-05,
+      "loss": 1.2694,
+      "step": 14030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9429572820663452,
+      "learning_rate": 4.629368226275443e-05,
+      "loss": 1.5489,
+      "step": 14031
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7363673448562622,
+      "learning_rate": 4.62934179222839e-05,
+      "loss": 1.4374,
+      "step": 14032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.823927879333496,
+      "learning_rate": 4.6293153581813376e-05,
+      "loss": 1.3493,
+      "step": 14033
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9597080945968628,
+      "learning_rate": 4.629288924134285e-05,
+      "loss": 1.4113,
+      "step": 14034
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.460224986076355,
+      "learning_rate": 4.629262490087233e-05,
+      "loss": 2.2051,
+      "step": 14035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3647596836090088,
+      "learning_rate": 4.62923605604018e-05,
+      "loss": 1.3127,
+      "step": 14036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8890752792358398,
+      "learning_rate": 4.629209621993128e-05,
+      "loss": 1.369,
+      "step": 14037
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7318083047866821,
+      "learning_rate": 4.6291831879460744e-05,
+      "loss": 1.3148,
+      "step": 14038
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.593575358390808,
+      "learning_rate": 4.6291567538990224e-05,
+      "loss": 1.4341,
+      "step": 14039
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.564911127090454,
+      "learning_rate": 4.62913031985197e-05,
+      "loss": 1.3128,
+      "step": 14040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.082202196121216,
+      "learning_rate": 4.629103885804917e-05,
+      "loss": 1.6578,
+      "step": 14041
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1911871433258057,
+      "learning_rate": 4.6290774517578645e-05,
+      "loss": 1.3062,
+      "step": 14042
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6923410892486572,
+      "learning_rate": 4.629051017710811e-05,
+      "loss": 1.3107,
+      "step": 14043
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8029969930648804,
+      "learning_rate": 4.629024583663759e-05,
+      "loss": 1.8345,
+      "step": 14044
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6876144409179688,
+      "learning_rate": 4.6289981496167066e-05,
+      "loss": 1.2663,
+      "step": 14045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.589981198310852,
+      "learning_rate": 4.628971715569654e-05,
+      "loss": 1.3616,
+      "step": 14046
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9194860458374023,
+      "learning_rate": 4.628945281522601e-05,
+      "loss": 1.4576,
+      "step": 14047
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.934843897819519,
+      "learning_rate": 4.6289188474755487e-05,
+      "loss": 1.3288,
+      "step": 14048
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.889999270439148,
+      "learning_rate": 4.628892413428496e-05,
+      "loss": 1.4065,
+      "step": 14049
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.218202590942383,
+      "learning_rate": 4.628865979381444e-05,
+      "loss": 1.3796,
+      "step": 14050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2772743701934814,
+      "learning_rate": 4.628839545334391e-05,
+      "loss": 1.4309,
+      "step": 14051
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8494335412979126,
+      "learning_rate": 4.628813111287338e-05,
+      "loss": 1.377,
+      "step": 14052
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.001399040222168,
+      "learning_rate": 4.6287866772402854e-05,
+      "loss": 1.4043,
+      "step": 14053
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1459708213806152,
+      "learning_rate": 4.628760243193233e-05,
+      "loss": 1.4156,
+      "step": 14054
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2189533710479736,
+      "learning_rate": 4.628733809146181e-05,
+      "loss": 1.493,
+      "step": 14055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7737116813659668,
+      "learning_rate": 4.6287073750991275e-05,
+      "loss": 1.2491,
+      "step": 14056
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1357061862945557,
+      "learning_rate": 4.6286809410520755e-05,
+      "loss": 1.5131,
+      "step": 14057
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6966081857681274,
+      "learning_rate": 4.628654507005022e-05,
+      "loss": 1.8642,
+      "step": 14058
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9163910150527954,
+      "learning_rate": 4.62862807295797e-05,
+      "loss": 1.5864,
+      "step": 14059
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.689073920249939,
+      "learning_rate": 4.6286016389109176e-05,
+      "loss": 1.4498,
+      "step": 14060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.924317717552185,
+      "learning_rate": 4.628575204863865e-05,
+      "loss": 1.4907,
+      "step": 14061
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.001495599746704,
+      "learning_rate": 4.628548770816812e-05,
+      "loss": 1.4413,
+      "step": 14062
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.035134792327881,
+      "learning_rate": 4.62852233676976e-05,
+      "loss": 1.5365,
+      "step": 14063
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9242562055587769,
+      "learning_rate": 4.628495902722707e-05,
+      "loss": 1.5067,
+      "step": 14064
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8317102193832397,
+      "learning_rate": 4.6284694686756544e-05,
+      "loss": 1.4706,
+      "step": 14065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.086989641189575,
+      "learning_rate": 4.628443034628602e-05,
+      "loss": 1.5178,
+      "step": 14066
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.594438314437866,
+      "learning_rate": 4.628416600581549e-05,
+      "loss": 1.5604,
+      "step": 14067
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6550482511520386,
+      "learning_rate": 4.6283901665344965e-05,
+      "loss": 1.4338,
+      "step": 14068
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9804189205169678,
+      "learning_rate": 4.628363732487444e-05,
+      "loss": 1.4561,
+      "step": 14069
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.886961579322815,
+      "learning_rate": 4.628337298440392e-05,
+      "loss": 1.5513,
+      "step": 14070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8348000049591064,
+      "learning_rate": 4.6283108643933385e-05,
+      "loss": 1.2929,
+      "step": 14071
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5620800256729126,
+      "learning_rate": 4.6282844303462866e-05,
+      "loss": 1.3703,
+      "step": 14072
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.828791618347168,
+      "learning_rate": 4.628257996299233e-05,
+      "loss": 1.5192,
+      "step": 14073
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5728532075881958,
+      "learning_rate": 4.628231562252181e-05,
+      "loss": 1.3561,
+      "step": 14074
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1646170616149902,
+      "learning_rate": 4.6282051282051287e-05,
+      "loss": 1.3896,
+      "step": 14075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5384725332260132,
+      "learning_rate": 4.628178694158076e-05,
+      "loss": 1.4844,
+      "step": 14076
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.494321584701538,
+      "learning_rate": 4.6281522601110234e-05,
+      "loss": 1.3493,
+      "step": 14077
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6504300832748413,
+      "learning_rate": 4.62812582606397e-05,
+      "loss": 1.4525,
+      "step": 14078
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.532557725906372,
+      "learning_rate": 4.628099392016918e-05,
+      "loss": 1.3796,
+      "step": 14079
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7849791049957275,
+      "learning_rate": 4.6280729579698654e-05,
+      "loss": 1.9457,
+      "step": 14080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7284764051437378,
+      "learning_rate": 4.628046523922813e-05,
+      "loss": 1.3227,
+      "step": 14081
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5715172290802,
+      "learning_rate": 4.62802008987576e-05,
+      "loss": 1.5466,
+      "step": 14082
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.975679636001587,
+      "learning_rate": 4.6279936558287075e-05,
+      "loss": 1.3851,
+      "step": 14083
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5927536487579346,
+      "learning_rate": 4.627967221781655e-05,
+      "loss": 1.3731,
+      "step": 14084
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.347548723220825,
+      "learning_rate": 4.627940787734603e-05,
+      "loss": 1.4625,
+      "step": 14085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.114037275314331,
+      "learning_rate": 4.6279143536875496e-05,
+      "loss": 1.4819,
+      "step": 14086
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7029473781585693,
+      "learning_rate": 4.6278879196404976e-05,
+      "loss": 1.3596,
+      "step": 14087
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.592419981956482,
+      "learning_rate": 4.627861485593444e-05,
+      "loss": 1.2822,
+      "step": 14088
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.593268632888794,
+      "learning_rate": 4.6278350515463917e-05,
+      "loss": 1.2959,
+      "step": 14089
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9656593799591064,
+      "learning_rate": 4.62780861749934e-05,
+      "loss": 1.3658,
+      "step": 14090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.740818738937378,
+      "learning_rate": 4.6277821834522864e-05,
+      "loss": 1.5302,
+      "step": 14091
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8608932495117188,
+      "learning_rate": 4.6277557494052344e-05,
+      "loss": 1.28,
+      "step": 14092
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9080768823623657,
+      "learning_rate": 4.627729315358181e-05,
+      "loss": 1.3968,
+      "step": 14093
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9527837038040161,
+      "learning_rate": 4.627702881311129e-05,
+      "loss": 1.5661,
+      "step": 14094
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.709858775138855,
+      "learning_rate": 4.6276764472640765e-05,
+      "loss": 1.3367,
+      "step": 14095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7446110248565674,
+      "learning_rate": 4.627650013217024e-05,
+      "loss": 1.447,
+      "step": 14096
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5533204078674316,
+      "learning_rate": 4.627623579169971e-05,
+      "loss": 1.2578,
+      "step": 14097
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.951812744140625,
+      "learning_rate": 4.6275971451229185e-05,
+      "loss": 1.4102,
+      "step": 14098
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.184286117553711,
+      "learning_rate": 4.627570711075866e-05,
+      "loss": 1.5953,
+      "step": 14099
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5261503458023071,
+      "learning_rate": 4.627544277028813e-05,
+      "loss": 1.3356,
+      "step": 14100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.399956464767456,
+      "learning_rate": 4.6275178429817606e-05,
+      "loss": 1.4045,
+      "step": 14101
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6087350845336914,
+      "learning_rate": 4.627491408934708e-05,
+      "loss": 1.5553,
+      "step": 14102
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4725781679153442,
+      "learning_rate": 4.627464974887655e-05,
+      "loss": 1.275,
+      "step": 14103
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3229289054870605,
+      "learning_rate": 4.627438540840603e-05,
+      "loss": 1.524,
+      "step": 14104
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.076655149459839,
+      "learning_rate": 4.627412106793551e-05,
+      "loss": 1.4473,
+      "step": 14105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.657784461975098,
+      "learning_rate": 4.6273856727464974e-05,
+      "loss": 1.4157,
+      "step": 14106
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.315361738204956,
+      "learning_rate": 4.6273592386994454e-05,
+      "loss": 1.5965,
+      "step": 14107
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.849949598312378,
+      "learning_rate": 4.627332804652392e-05,
+      "loss": 1.2958,
+      "step": 14108
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3905434608459473,
+      "learning_rate": 4.62730637060534e-05,
+      "loss": 1.5608,
+      "step": 14109
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7481305599212646,
+      "learning_rate": 4.6272799365582875e-05,
+      "loss": 1.4262,
+      "step": 14110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3123449087142944,
+      "learning_rate": 4.627253502511235e-05,
+      "loss": 1.253,
+      "step": 14111
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2868669033050537,
+      "learning_rate": 4.627227068464182e-05,
+      "loss": 1.5886,
+      "step": 14112
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5204181671142578,
+      "learning_rate": 4.6272006344171296e-05,
+      "loss": 1.2539,
+      "step": 14113
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.507594347000122,
+      "learning_rate": 4.627174200370077e-05,
+      "loss": 1.4344,
+      "step": 14114
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8537089824676514,
+      "learning_rate": 4.627147766323024e-05,
+      "loss": 1.4522,
+      "step": 14115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6285532712936401,
+      "learning_rate": 4.6271213322759716e-05,
+      "loss": 1.3812,
+      "step": 14116
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6984553337097168,
+      "learning_rate": 4.627094898228919e-05,
+      "loss": 1.2463,
+      "step": 14117
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.302188754081726,
+      "learning_rate": 4.6270684641818664e-05,
+      "loss": 1.22,
+      "step": 14118
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6592952013015747,
+      "learning_rate": 4.627042030134814e-05,
+      "loss": 1.479,
+      "step": 14119
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9380676746368408,
+      "learning_rate": 4.627015596087762e-05,
+      "loss": 1.512,
+      "step": 14120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8038333654403687,
+      "learning_rate": 4.6269891620407084e-05,
+      "loss": 2.4637,
+      "step": 14121
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8182132244110107,
+      "learning_rate": 4.6269627279936565e-05,
+      "loss": 1.5004,
+      "step": 14122
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7096428871154785,
+      "learning_rate": 4.626936293946603e-05,
+      "loss": 1.3507,
+      "step": 14123
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9131666421890259,
+      "learning_rate": 4.6269098598995505e-05,
+      "loss": 1.3356,
+      "step": 14124
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6760681867599487,
+      "learning_rate": 4.6268834258524985e-05,
+      "loss": 1.3053,
+      "step": 14125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8435406684875488,
+      "learning_rate": 4.626856991805445e-05,
+      "loss": 2.3039,
+      "step": 14126
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9951635599136353,
+      "learning_rate": 4.626830557758393e-05,
+      "loss": 1.3512,
+      "step": 14127
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6684437990188599,
+      "learning_rate": 4.62680412371134e-05,
+      "loss": 1.3703,
+      "step": 14128
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7078635692596436,
+      "learning_rate": 4.626777689664288e-05,
+      "loss": 1.3772,
+      "step": 14129
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8003606796264648,
+      "learning_rate": 4.626751255617235e-05,
+      "loss": 1.3797,
+      "step": 14130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5514687299728394,
+      "learning_rate": 4.626724821570183e-05,
+      "loss": 1.4576,
+      "step": 14131
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.268419027328491,
+      "learning_rate": 4.62669838752313e-05,
+      "loss": 1.3829,
+      "step": 14132
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.721902847290039,
+      "learning_rate": 4.6266719534760774e-05,
+      "loss": 1.2402,
+      "step": 14133
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.83120596408844,
+      "learning_rate": 4.626645519429025e-05,
+      "loss": 1.3947,
+      "step": 14134
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4843075275421143,
+      "learning_rate": 4.626619085381972e-05,
+      "loss": 2.184,
+      "step": 14135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.678452730178833,
+      "learning_rate": 4.6265926513349195e-05,
+      "loss": 1.4794,
+      "step": 14136
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.012140989303589,
+      "learning_rate": 4.626566217287867e-05,
+      "loss": 1.5912,
+      "step": 14137
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3703334331512451,
+      "learning_rate": 4.626539783240814e-05,
+      "loss": 1.388,
+      "step": 14138
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4245940446853638,
+      "learning_rate": 4.6265133491937615e-05,
+      "loss": 1.3255,
+      "step": 14139
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7916029691696167,
+      "learning_rate": 4.6264869151467096e-05,
+      "loss": 1.4523,
+      "step": 14140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7517338991165161,
+      "learning_rate": 4.626460481099656e-05,
+      "loss": 1.3115,
+      "step": 14141
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8345541954040527,
+      "learning_rate": 4.626434047052604e-05,
+      "loss": 1.4572,
+      "step": 14142
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2619330883026123,
+      "learning_rate": 4.626407613005551e-05,
+      "loss": 1.4611,
+      "step": 14143
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9531422853469849,
+      "learning_rate": 4.626381178958499e-05,
+      "loss": 1.3351,
+      "step": 14144
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5813109874725342,
+      "learning_rate": 4.6263547449114464e-05,
+      "loss": 1.1151,
+      "step": 14145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8847966194152832,
+      "learning_rate": 4.626328310864394e-05,
+      "loss": 1.486,
+      "step": 14146
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5029983520507812,
+      "learning_rate": 4.626301876817341e-05,
+      "loss": 1.2862,
+      "step": 14147
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6617956161499023,
+      "learning_rate": 4.6262754427702884e-05,
+      "loss": 1.3438,
+      "step": 14148
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7136305570602417,
+      "learning_rate": 4.626249008723236e-05,
+      "loss": 1.3316,
+      "step": 14149
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9067291021347046,
+      "learning_rate": 4.626222574676183e-05,
+      "loss": 1.3167,
+      "step": 14150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7564877271652222,
+      "learning_rate": 4.6261961406291305e-05,
+      "loss": 1.4399,
+      "step": 14151
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7490618228912354,
+      "learning_rate": 4.626169706582078e-05,
+      "loss": 1.432,
+      "step": 14152
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.080268383026123,
+      "learning_rate": 4.626143272535025e-05,
+      "loss": 1.4552,
+      "step": 14153
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2022438049316406,
+      "learning_rate": 4.6261168384879726e-05,
+      "loss": 1.4752,
+      "step": 14154
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7848877906799316,
+      "learning_rate": 4.6260904044409206e-05,
+      "loss": 1.4475,
+      "step": 14155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0189967155456543,
+      "learning_rate": 4.626063970393867e-05,
+      "loss": 1.4349,
+      "step": 14156
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5329008102416992,
+      "learning_rate": 4.626037536346815e-05,
+      "loss": 1.4389,
+      "step": 14157
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6568962335586548,
+      "learning_rate": 4.626011102299762e-05,
+      "loss": 1.3978,
+      "step": 14158
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6664414405822754,
+      "learning_rate": 4.62598466825271e-05,
+      "loss": 2.1759,
+      "step": 14159
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6500329971313477,
+      "learning_rate": 4.6259582342056574e-05,
+      "loss": 1.4985,
+      "step": 14160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6603220701217651,
+      "learning_rate": 4.625931800158604e-05,
+      "loss": 1.4426,
+      "step": 14161
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0328550338745117,
+      "learning_rate": 4.625905366111552e-05,
+      "loss": 1.3723,
+      "step": 14162
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8852412700653076,
+      "learning_rate": 4.625878932064499e-05,
+      "loss": 1.3816,
+      "step": 14163
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.255490779876709,
+      "learning_rate": 4.625852498017447e-05,
+      "loss": 1.538,
+      "step": 14164
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9066532850265503,
+      "learning_rate": 4.625826063970394e-05,
+      "loss": 1.5246,
+      "step": 14165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.806774377822876,
+      "learning_rate": 4.6257996299233415e-05,
+      "loss": 1.368,
+      "step": 14166
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7820814847946167,
+      "learning_rate": 4.625773195876289e-05,
+      "loss": 1.457,
+      "step": 14167
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9580113887786865,
+      "learning_rate": 4.625746761829236e-05,
+      "loss": 1.4251,
+      "step": 14168
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6477724313735962,
+      "learning_rate": 4.6257203277821836e-05,
+      "loss": 1.3237,
+      "step": 14169
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6929701566696167,
+      "learning_rate": 4.625693893735131e-05,
+      "loss": 1.5001,
+      "step": 14170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7715197801589966,
+      "learning_rate": 4.625667459688078e-05,
+      "loss": 1.3829,
+      "step": 14171
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4336285591125488,
+      "learning_rate": 4.625641025641026e-05,
+      "loss": 1.2215,
+      "step": 14172
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5594063997268677,
+      "learning_rate": 4.625614591593973e-05,
+      "loss": 1.2308,
+      "step": 14173
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9070755243301392,
+      "learning_rate": 4.6255881575469204e-05,
+      "loss": 1.6518,
+      "step": 14174
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8345634937286377,
+      "learning_rate": 4.6255617234998684e-05,
+      "loss": 1.4733,
+      "step": 14175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5825974941253662,
+      "learning_rate": 4.625535289452815e-05,
+      "loss": 1.403,
+      "step": 14176
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6930991411209106,
+      "learning_rate": 4.625508855405763e-05,
+      "loss": 1.3215,
+      "step": 14177
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9421513080596924,
+      "learning_rate": 4.62548242135871e-05,
+      "loss": 1.4863,
+      "step": 14178
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5362321138381958,
+      "learning_rate": 4.625455987311658e-05,
+      "loss": 1.2678,
+      "step": 14179
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4533145427703857,
+      "learning_rate": 4.625429553264605e-05,
+      "loss": 1.6584,
+      "step": 14180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7325406074523926,
+      "learning_rate": 4.6254031192175526e-05,
+      "loss": 1.3501,
+      "step": 14181
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0828957557678223,
+      "learning_rate": 4.6253766851705e-05,
+      "loss": 1.4901,
+      "step": 14182
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5821731090545654,
+      "learning_rate": 4.625350251123447e-05,
+      "loss": 1.4009,
+      "step": 14183
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6264276504516602,
+      "learning_rate": 4.6253238170763946e-05,
+      "loss": 1.3237,
+      "step": 14184
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.752625823020935,
+      "learning_rate": 4.625297383029342e-05,
+      "loss": 1.2227,
+      "step": 14185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4713459014892578,
+      "learning_rate": 4.6252709489822894e-05,
+      "loss": 1.1968,
+      "step": 14186
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9904353618621826,
+      "learning_rate": 4.625244514935237e-05,
+      "loss": 1.3759,
+      "step": 14187
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7224763631820679,
+      "learning_rate": 4.625218080888184e-05,
+      "loss": 1.3751,
+      "step": 14188
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6406586170196533,
+      "learning_rate": 4.6251916468411314e-05,
+      "loss": 1.8201,
+      "step": 14189
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.037283420562744,
+      "learning_rate": 4.6251652127940795e-05,
+      "loss": 1.3436,
+      "step": 14190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6509361267089844,
+      "learning_rate": 4.625138778747026e-05,
+      "loss": 1.3168,
+      "step": 14191
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7333030700683594,
+      "learning_rate": 4.625112344699974e-05,
+      "loss": 1.4153,
+      "step": 14192
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6356918811798096,
+      "learning_rate": 4.625085910652921e-05,
+      "loss": 1.3791,
+      "step": 14193
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6231790781021118,
+      "learning_rate": 4.625059476605869e-05,
+      "loss": 1.4545,
+      "step": 14194
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8454104661941528,
+      "learning_rate": 4.625033042558816e-05,
+      "loss": 1.3119,
+      "step": 14195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5849887132644653,
+      "learning_rate": 4.625006608511763e-05,
+      "loss": 1.2887,
+      "step": 14196
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.618725299835205,
+      "learning_rate": 4.624980174464711e-05,
+      "loss": 1.3285,
+      "step": 14197
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6287063360214233,
+      "learning_rate": 4.6249537404176576e-05,
+      "loss": 1.4251,
+      "step": 14198
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5340657234191895,
+      "learning_rate": 4.624927306370606e-05,
+      "loss": 1.3512,
+      "step": 14199
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1311609745025635,
+      "learning_rate": 4.624900872323553e-05,
+      "loss": 1.3147,
+      "step": 14200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9374690055847168,
+      "learning_rate": 4.6248744382765004e-05,
+      "loss": 1.2826,
+      "step": 14201
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8847157955169678,
+      "learning_rate": 4.624848004229448e-05,
+      "loss": 1.3285,
+      "step": 14202
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6264277696609497,
+      "learning_rate": 4.624821570182395e-05,
+      "loss": 1.2533,
+      "step": 14203
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4306031465530396,
+      "learning_rate": 4.6247951361353425e-05,
+      "loss": 1.3545,
+      "step": 14204
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.652656316757202,
+      "learning_rate": 4.6247687020882905e-05,
+      "loss": 1.4496,
+      "step": 14205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0605783462524414,
+      "learning_rate": 4.624742268041237e-05,
+      "loss": 1.505,
+      "step": 14206
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5887857675552368,
+      "learning_rate": 4.6247158339941845e-05,
+      "loss": 1.2476,
+      "step": 14207
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7384761571884155,
+      "learning_rate": 4.624689399947132e-05,
+      "loss": 1.3024,
+      "step": 14208
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.162607192993164,
+      "learning_rate": 4.624662965900079e-05,
+      "loss": 1.41,
+      "step": 14209
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1566355228424072,
+      "learning_rate": 4.624636531853027e-05,
+      "loss": 1.5053,
+      "step": 14210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.603599190711975,
+      "learning_rate": 4.624610097805974e-05,
+      "loss": 1.3746,
+      "step": 14211
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6048988103866577,
+      "learning_rate": 4.624583663758922e-05,
+      "loss": 1.3835,
+      "step": 14212
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.414908409118652,
+      "learning_rate": 4.624557229711869e-05,
+      "loss": 1.4826,
+      "step": 14213
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6094646453857422,
+      "learning_rate": 4.624530795664817e-05,
+      "loss": 1.3914,
+      "step": 14214
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8879809379577637,
+      "learning_rate": 4.624504361617764e-05,
+      "loss": 1.1742,
+      "step": 14215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.110144853591919,
+      "learning_rate": 4.6244779275707114e-05,
+      "loss": 1.4491,
+      "step": 14216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8500874042510986,
+      "learning_rate": 4.624451493523659e-05,
+      "loss": 1.444,
+      "step": 14217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.630568265914917,
+      "learning_rate": 4.624425059476606e-05,
+      "loss": 1.3079,
+      "step": 14218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.729778528213501,
+      "learning_rate": 4.6243986254295535e-05,
+      "loss": 1.2905,
+      "step": 14219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9288601875305176,
+      "learning_rate": 4.624372191382501e-05,
+      "loss": 1.4701,
+      "step": 14220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0972585678100586,
+      "learning_rate": 4.624345757335448e-05,
+      "loss": 1.5008,
+      "step": 14221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.593290328979492,
+      "learning_rate": 4.6243193232883956e-05,
+      "loss": 1.5157,
+      "step": 14222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.764395833015442,
+      "learning_rate": 4.624292889241343e-05,
+      "loss": 1.3096,
+      "step": 14223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2933427095413208,
+      "learning_rate": 4.62426645519429e-05,
+      "loss": 1.1936,
+      "step": 14224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7049107551574707,
+      "learning_rate": 4.624240021147238e-05,
+      "loss": 1.3625,
+      "step": 14225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.006927967071533,
+      "learning_rate": 4.624213587100185e-05,
+      "loss": 1.3833,
+      "step": 14226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4969048500061035,
+      "learning_rate": 4.624187153053133e-05,
+      "loss": 1.523,
+      "step": 14227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4613687992095947,
+      "learning_rate": 4.62416071900608e-05,
+      "loss": 1.2725,
+      "step": 14228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.524609923362732,
+      "learning_rate": 4.624134284959028e-05,
+      "loss": 1.4116,
+      "step": 14229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0986409187316895,
+      "learning_rate": 4.624107850911975e-05,
+      "loss": 1.4634,
+      "step": 14230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2627012729644775,
+      "learning_rate": 4.6240814168649225e-05,
+      "loss": 1.448,
+      "step": 14231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9963035583496094,
+      "learning_rate": 4.62405498281787e-05,
+      "loss": 1.4722,
+      "step": 14232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5466513633728027,
+      "learning_rate": 4.6240285487708165e-05,
+      "loss": 1.3715,
+      "step": 14233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.968032121658325,
+      "learning_rate": 4.6240021147237645e-05,
+      "loss": 1.6059,
+      "step": 14234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5692949295043945,
+      "learning_rate": 4.623975680676712e-05,
+      "loss": 1.2233,
+      "step": 14235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8453270196914673,
+      "learning_rate": 4.623949246629659e-05,
+      "loss": 1.251,
+      "step": 14236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5815695524215698,
+      "learning_rate": 4.6239228125826066e-05,
+      "loss": 1.416,
+      "step": 14237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6598219871520996,
+      "learning_rate": 4.623896378535554e-05,
+      "loss": 1.4412,
+      "step": 14238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8152031898498535,
+      "learning_rate": 4.623869944488501e-05,
+      "loss": 1.4252,
+      "step": 14239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.775070309638977,
+      "learning_rate": 4.6238435104414493e-05,
+      "loss": 1.4198,
+      "step": 14240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.496087908744812,
+      "learning_rate": 4.623817076394396e-05,
+      "loss": 1.7692,
+      "step": 14241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1075422763824463,
+      "learning_rate": 4.6237906423473434e-05,
+      "loss": 1.4545,
+      "step": 14242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9688948392868042,
+      "learning_rate": 4.623764208300291e-05,
+      "loss": 1.4021,
+      "step": 14243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7926928997039795,
+      "learning_rate": 4.623737774253238e-05,
+      "loss": 1.33,
+      "step": 14244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.77274489402771,
+      "learning_rate": 4.623711340206186e-05,
+      "loss": 1.328,
+      "step": 14245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6869959831237793,
+      "learning_rate": 4.623684906159133e-05,
+      "loss": 1.3285,
+      "step": 14246
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5949838161468506,
+      "learning_rate": 4.623658472112081e-05,
+      "loss": 1.3517,
+      "step": 14247
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0282528400421143,
+      "learning_rate": 4.6236320380650275e-05,
+      "loss": 2.3643,
+      "step": 14248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6111869812011719,
+      "learning_rate": 4.6236056040179756e-05,
+      "loss": 1.3578,
+      "step": 14249
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.663125991821289,
+      "learning_rate": 4.623579169970923e-05,
+      "loss": 1.3637,
+      "step": 14250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9850473403930664,
+      "learning_rate": 4.62355273592387e-05,
+      "loss": 1.3495,
+      "step": 14251
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5679409503936768,
+      "learning_rate": 4.6235263018768176e-05,
+      "loss": 1.4913,
+      "step": 14252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3655123710632324,
+      "learning_rate": 4.623499867829765e-05,
+      "loss": 1.5481,
+      "step": 14253
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9886871576309204,
+      "learning_rate": 4.6234734337827123e-05,
+      "loss": 1.4998,
+      "step": 14254
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6732151508331299,
+      "learning_rate": 4.62344699973566e-05,
+      "loss": 1.4341,
+      "step": 14255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9663739204406738,
+      "learning_rate": 4.623420565688607e-05,
+      "loss": 1.3559,
+      "step": 14256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8975305557250977,
+      "learning_rate": 4.6233941316415544e-05,
+      "loss": 1.4617,
+      "step": 14257
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.091631889343262,
+      "learning_rate": 4.623367697594502e-05,
+      "loss": 1.9803,
+      "step": 14258
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2321882247924805,
+      "learning_rate": 4.623341263547449e-05,
+      "loss": 1.2125,
+      "step": 14259
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5800589323043823,
+      "learning_rate": 4.623314829500397e-05,
+      "loss": 1.296,
+      "step": 14260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5042340755462646,
+      "learning_rate": 4.623288395453344e-05,
+      "loss": 1.2632,
+      "step": 14261
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8782007694244385,
+      "learning_rate": 4.623261961406292e-05,
+      "loss": 1.3786,
+      "step": 14262
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1570446491241455,
+      "learning_rate": 4.6232355273592386e-05,
+      "loss": 1.404,
+      "step": 14263
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.789101243019104,
+      "learning_rate": 4.6232090933121866e-05,
+      "loss": 1.4045,
+      "step": 14264
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9024838209152222,
+      "learning_rate": 4.623182659265134e-05,
+      "loss": 1.4069,
+      "step": 14265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9318182468414307,
+      "learning_rate": 4.623156225218081e-05,
+      "loss": 1.2919,
+      "step": 14266
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7589616775512695,
+      "learning_rate": 4.623129791171029e-05,
+      "loss": 1.4577,
+      "step": 14267
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5188945531845093,
+      "learning_rate": 4.623103357123976e-05,
+      "loss": 1.2958,
+      "step": 14268
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6669365167617798,
+      "learning_rate": 4.6230769230769234e-05,
+      "loss": 1.344,
+      "step": 14269
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3128113746643066,
+      "learning_rate": 4.623050489029871e-05,
+      "loss": 1.4233,
+      "step": 14270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8598544597625732,
+      "learning_rate": 4.623024054982818e-05,
+      "loss": 1.5331,
+      "step": 14271
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9192848205566406,
+      "learning_rate": 4.6229976209357655e-05,
+      "loss": 1.3187,
+      "step": 14272
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.764348030090332,
+      "learning_rate": 4.622971186888713e-05,
+      "loss": 1.3684,
+      "step": 14273
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0475189685821533,
+      "learning_rate": 4.62294475284166e-05,
+      "loss": 1.4348,
+      "step": 14274
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8364074230194092,
+      "learning_rate": 4.622918318794608e-05,
+      "loss": 1.4262,
+      "step": 14275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4690407514572144,
+      "learning_rate": 4.622891884747555e-05,
+      "loss": 1.4157,
+      "step": 14276
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.340861201286316,
+      "learning_rate": 4.622865450700503e-05,
+      "loss": 1.1984,
+      "step": 14277
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6128664016723633,
+      "learning_rate": 4.6228390166534496e-05,
+      "loss": 1.4124,
+      "step": 14278
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7781715393066406,
+      "learning_rate": 4.622812582606397e-05,
+      "loss": 1.3673,
+      "step": 14279
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.311187505722046,
+      "learning_rate": 4.622786148559345e-05,
+      "loss": 1.5764,
+      "step": 14280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0230956077575684,
+      "learning_rate": 4.622759714512292e-05,
+      "loss": 1.3856,
+      "step": 14281
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4903424978256226,
+      "learning_rate": 4.62273328046524e-05,
+      "loss": 1.3661,
+      "step": 14282
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3395036458969116,
+      "learning_rate": 4.6227068464181864e-05,
+      "loss": 1.2335,
+      "step": 14283
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5160939693450928,
+      "learning_rate": 4.6226804123711344e-05,
+      "loss": 1.2854,
+      "step": 14284
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5413930416107178,
+      "learning_rate": 4.622653978324082e-05,
+      "loss": 1.5318,
+      "step": 14285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5384167432785034,
+      "learning_rate": 4.622627544277029e-05,
+      "loss": 1.2577,
+      "step": 14286
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5475432872772217,
+      "learning_rate": 4.6226011102299765e-05,
+      "loss": 1.2581,
+      "step": 14287
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.947843074798584,
+      "learning_rate": 4.622574676182924e-05,
+      "loss": 1.4082,
+      "step": 14288
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5896092653274536,
+      "learning_rate": 4.622548242135871e-05,
+      "loss": 1.2998,
+      "step": 14289
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.455345630645752,
+      "learning_rate": 4.6225218080888186e-05,
+      "loss": 1.9569,
+      "step": 14290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6176962852478027,
+      "learning_rate": 4.622495374041766e-05,
+      "loss": 1.3226,
+      "step": 14291
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4363994598388672,
+      "learning_rate": 4.622468939994713e-05,
+      "loss": 1.2559,
+      "step": 14292
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6751904487609863,
+      "learning_rate": 4.6224425059476606e-05,
+      "loss": 1.4123,
+      "step": 14293
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8766719102859497,
+      "learning_rate": 4.622416071900608e-05,
+      "loss": 1.5611,
+      "step": 14294
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7729477882385254,
+      "learning_rate": 4.622389637853556e-05,
+      "loss": 1.3298,
+      "step": 14295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0429635047912598,
+      "learning_rate": 4.622363203806503e-05,
+      "loss": 1.4131,
+      "step": 14296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0133156776428223,
+      "learning_rate": 4.622336769759451e-05,
+      "loss": 1.5803,
+      "step": 14297
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0795862674713135,
+      "learning_rate": 4.6223103357123974e-05,
+      "loss": 1.403,
+      "step": 14298
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.43587327003479,
+      "learning_rate": 4.6222839016653454e-05,
+      "loss": 1.5698,
+      "step": 14299
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.962512493133545,
+      "learning_rate": 4.622257467618293e-05,
+      "loss": 1.4039,
+      "step": 14300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8463313579559326,
+      "learning_rate": 4.62223103357124e-05,
+      "loss": 1.3488,
+      "step": 14301
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5230066776275635,
+      "learning_rate": 4.6222045995241875e-05,
+      "loss": 1.2588,
+      "step": 14302
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9380455017089844,
+      "learning_rate": 4.622178165477135e-05,
+      "loss": 1.3175,
+      "step": 14303
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6259331703186035,
+      "learning_rate": 4.622151731430082e-05,
+      "loss": 1.4964,
+      "step": 14304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8423740863800049,
+      "learning_rate": 4.6221252973830296e-05,
+      "loss": 1.4353,
+      "step": 14305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7133678197860718,
+      "learning_rate": 4.622098863335977e-05,
+      "loss": 1.3545,
+      "step": 14306
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9834249019622803,
+      "learning_rate": 4.622072429288924e-05,
+      "loss": 1.4588,
+      "step": 14307
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8788880109786987,
+      "learning_rate": 4.622045995241872e-05,
+      "loss": 1.4344,
+      "step": 14308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.251574754714966,
+      "learning_rate": 4.622019561194819e-05,
+      "loss": 1.4242,
+      "step": 14309
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2358827590942383,
+      "learning_rate": 4.621993127147767e-05,
+      "loss": 1.5382,
+      "step": 14310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4082260131835938,
+      "learning_rate": 4.621966693100714e-05,
+      "loss": 1.3808,
+      "step": 14311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.900610089302063,
+      "learning_rate": 4.621940259053662e-05,
+      "loss": 1.3727,
+      "step": 14312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5907410383224487,
+      "learning_rate": 4.6219138250066084e-05,
+      "loss": 1.2787,
+      "step": 14313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7147927284240723,
+      "learning_rate": 4.621887390959556e-05,
+      "loss": 1.4085,
+      "step": 14314
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9568370580673218,
+      "learning_rate": 4.621860956912504e-05,
+      "loss": 1.3649,
+      "step": 14315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4867316484451294,
+      "learning_rate": 4.6218345228654505e-05,
+      "loss": 1.6902,
+      "step": 14316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.767515778541565,
+      "learning_rate": 4.6218080888183986e-05,
+      "loss": 1.4697,
+      "step": 14317
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9848283529281616,
+      "learning_rate": 4.621781654771345e-05,
+      "loss": 1.3498,
+      "step": 14318
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6899150609970093,
+      "learning_rate": 4.621755220724293e-05,
+      "loss": 1.2998,
+      "step": 14319
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6074731349945068,
+      "learning_rate": 4.6217287866772406e-05,
+      "loss": 1.2706,
+      "step": 14320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.493303060531616,
+      "learning_rate": 4.621702352630188e-05,
+      "loss": 2.1443,
+      "step": 14321
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5084384679794312,
+      "learning_rate": 4.6216759185831353e-05,
+      "loss": 1.2892,
+      "step": 14322
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4212322235107422,
+      "learning_rate": 4.621649484536083e-05,
+      "loss": 1.1689,
+      "step": 14323
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7125345468521118,
+      "learning_rate": 4.62162305048903e-05,
+      "loss": 1.3146,
+      "step": 14324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.912008285522461,
+      "learning_rate": 4.6215966164419774e-05,
+      "loss": 1.4689,
+      "step": 14325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6811636686325073,
+      "learning_rate": 4.621570182394925e-05,
+      "loss": 1.2789,
+      "step": 14326
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.07745099067688,
+      "learning_rate": 4.621543748347872e-05,
+      "loss": 1.3437,
+      "step": 14327
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7015630006790161,
+      "learning_rate": 4.6215173143008195e-05,
+      "loss": 1.323,
+      "step": 14328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0455050468444824,
+      "learning_rate": 4.621490880253767e-05,
+      "loss": 1.2889,
+      "step": 14329
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4521759748458862,
+      "learning_rate": 4.621464446206715e-05,
+      "loss": 1.3247,
+      "step": 14330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7726022005081177,
+      "learning_rate": 4.6214380121596616e-05,
+      "loss": 1.3273,
+      "step": 14331
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8682966232299805,
+      "learning_rate": 4.6214115781126096e-05,
+      "loss": 1.496,
+      "step": 14332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.086334466934204,
+      "learning_rate": 4.621385144065556e-05,
+      "loss": 1.4347,
+      "step": 14333
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9720834493637085,
+      "learning_rate": 4.621358710018504e-05,
+      "loss": 1.4741,
+      "step": 14334
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.869608163833618,
+      "learning_rate": 4.6213322759714517e-05,
+      "loss": 1.5768,
+      "step": 14335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9717278480529785,
+      "learning_rate": 4.621305841924399e-05,
+      "loss": 1.4411,
+      "step": 14336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0693371295928955,
+      "learning_rate": 4.6212794078773464e-05,
+      "loss": 1.3652,
+      "step": 14337
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6011420488357544,
+      "learning_rate": 4.621252973830294e-05,
+      "loss": 1.3076,
+      "step": 14338
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.632907748222351,
+      "learning_rate": 4.621226539783241e-05,
+      "loss": 1.2868,
+      "step": 14339
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5055246353149414,
+      "learning_rate": 4.6212001057361884e-05,
+      "loss": 1.1922,
+      "step": 14340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9063732624053955,
+      "learning_rate": 4.621173671689136e-05,
+      "loss": 1.6952,
+      "step": 14341
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0545544624328613,
+      "learning_rate": 4.621147237642083e-05,
+      "loss": 1.4798,
+      "step": 14342
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8115735054016113,
+      "learning_rate": 4.6211208035950305e-05,
+      "loss": 1.2717,
+      "step": 14343
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.805338740348816,
+      "learning_rate": 4.621094369547978e-05,
+      "loss": 1.3846,
+      "step": 14344
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5480430126190186,
+      "learning_rate": 4.621067935500926e-05,
+      "loss": 1.2439,
+      "step": 14345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9159806966781616,
+      "learning_rate": 4.6210415014538726e-05,
+      "loss": 1.3996,
+      "step": 14346
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9632039070129395,
+      "learning_rate": 4.6210150674068206e-05,
+      "loss": 1.3437,
+      "step": 14347
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8598742485046387,
+      "learning_rate": 4.620988633359767e-05,
+      "loss": 1.2452,
+      "step": 14348
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9036879539489746,
+      "learning_rate": 4.620962199312715e-05,
+      "loss": 1.56,
+      "step": 14349
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0618677139282227,
+      "learning_rate": 4.620935765265663e-05,
+      "loss": 1.4481,
+      "step": 14350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7236939668655396,
+      "learning_rate": 4.6209093312186094e-05,
+      "loss": 1.4547,
+      "step": 14351
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.698969841003418,
+      "learning_rate": 4.6208828971715574e-05,
+      "loss": 1.4904,
+      "step": 14352
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.528774380683899,
+      "learning_rate": 4.620856463124504e-05,
+      "loss": 1.3479,
+      "step": 14353
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.77311110496521,
+      "learning_rate": 4.620830029077452e-05,
+      "loss": 1.2787,
+      "step": 14354
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0783402919769287,
+      "learning_rate": 4.6208035950303995e-05,
+      "loss": 1.3079,
+      "step": 14355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8295838832855225,
+      "learning_rate": 4.620777160983347e-05,
+      "loss": 1.3507,
+      "step": 14356
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.733041524887085,
+      "learning_rate": 4.620750726936294e-05,
+      "loss": 1.291,
+      "step": 14357
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8480783700942993,
+      "learning_rate": 4.6207242928892416e-05,
+      "loss": 1.4018,
+      "step": 14358
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.150249719619751,
+      "learning_rate": 4.620697858842189e-05,
+      "loss": 1.5789,
+      "step": 14359
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6335296630859375,
+      "learning_rate": 4.620671424795136e-05,
+      "loss": 1.3156,
+      "step": 14360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6247615814208984,
+      "learning_rate": 4.6206449907480836e-05,
+      "loss": 1.4362,
+      "step": 14361
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5736265182495117,
+      "learning_rate": 4.620618556701031e-05,
+      "loss": 1.3811,
+      "step": 14362
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7452802658081055,
+      "learning_rate": 4.620592122653978e-05,
+      "loss": 1.4472,
+      "step": 14363
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8345677852630615,
+      "learning_rate": 4.620565688606926e-05,
+      "loss": 1.4496,
+      "step": 14364
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8593333959579468,
+      "learning_rate": 4.620539254559874e-05,
+      "loss": 1.2881,
+      "step": 14365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7150501012802124,
+      "learning_rate": 4.6205128205128204e-05,
+      "loss": 2.0723,
+      "step": 14366
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7188664674758911,
+      "learning_rate": 4.6204863864657684e-05,
+      "loss": 1.4018,
+      "step": 14367
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7766218185424805,
+      "learning_rate": 4.620459952418715e-05,
+      "loss": 1.472,
+      "step": 14368
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.950869083404541,
+      "learning_rate": 4.620433518371663e-05,
+      "loss": 1.3724,
+      "step": 14369
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4383723735809326,
+      "learning_rate": 4.6204070843246105e-05,
+      "loss": 1.5133,
+      "step": 14370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2546911239624023,
+      "learning_rate": 4.620380650277558e-05,
+      "loss": 1.4982,
+      "step": 14371
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0565946102142334,
+      "learning_rate": 4.620354216230505e-05,
+      "loss": 1.2275,
+      "step": 14372
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.699897289276123,
+      "learning_rate": 4.6203277821834526e-05,
+      "loss": 1.2139,
+      "step": 14373
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5756715536117554,
+      "learning_rate": 4.6203013481364e-05,
+      "loss": 1.2646,
+      "step": 14374
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4458073377609253,
+      "learning_rate": 4.620274914089347e-05,
+      "loss": 1.294,
+      "step": 14375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.795432686805725,
+      "learning_rate": 4.6202484800422947e-05,
+      "loss": 1.4256,
+      "step": 14376
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7938947677612305,
+      "learning_rate": 4.620222045995242e-05,
+      "loss": 1.4353,
+      "step": 14377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4702117443084717,
+      "learning_rate": 4.6201956119481894e-05,
+      "loss": 1.4262,
+      "step": 14378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8744440078735352,
+      "learning_rate": 4.620169177901137e-05,
+      "loss": 1.4261,
+      "step": 14379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8728129863739014,
+      "learning_rate": 4.620142743854085e-05,
+      "loss": 1.4464,
+      "step": 14380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6687089204788208,
+      "learning_rate": 4.6201163098070314e-05,
+      "loss": 1.3876,
+      "step": 14381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4637393951416016,
+      "learning_rate": 4.6200898757599795e-05,
+      "loss": 1.5433,
+      "step": 14382
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6879361867904663,
+      "learning_rate": 4.620063441712926e-05,
+      "loss": 1.3945,
+      "step": 14383
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.623437762260437,
+      "learning_rate": 4.620037007665874e-05,
+      "loss": 1.3642,
+      "step": 14384
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.242344379425049,
+      "learning_rate": 4.6200105736188215e-05,
+      "loss": 1.4581,
+      "step": 14385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3847997188568115,
+      "learning_rate": 4.619984139571769e-05,
+      "loss": 1.3453,
+      "step": 14386
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.62337064743042,
+      "learning_rate": 4.619957705524716e-05,
+      "loss": 1.3654,
+      "step": 14387
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1965560913085938,
+      "learning_rate": 4.619931271477663e-05,
+      "loss": 1.4986,
+      "step": 14388
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.598210573196411,
+      "learning_rate": 4.619904837430611e-05,
+      "loss": 1.5427,
+      "step": 14389
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0088937282562256,
+      "learning_rate": 4.619878403383558e-05,
+      "loss": 1.4536,
+      "step": 14390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.827500820159912,
+      "learning_rate": 4.619851969336506e-05,
+      "loss": 1.462,
+      "step": 14391
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8823267221450806,
+      "learning_rate": 4.619825535289453e-05,
+      "loss": 1.4611,
+      "step": 14392
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.883378505706787,
+      "learning_rate": 4.6197991012424004e-05,
+      "loss": 1.3434,
+      "step": 14393
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5878115892410278,
+      "learning_rate": 4.619772667195348e-05,
+      "loss": 1.3461,
+      "step": 14394
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6200100183486938,
+      "learning_rate": 4.619746233148296e-05,
+      "loss": 1.2686,
+      "step": 14395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.292302131652832,
+      "learning_rate": 4.6197197991012425e-05,
+      "loss": 1.2761,
+      "step": 14396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.685063362121582,
+      "learning_rate": 4.61969336505419e-05,
+      "loss": 1.267,
+      "step": 14397
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.758954644203186,
+      "learning_rate": 4.619666931007137e-05,
+      "loss": 1.465,
+      "step": 14398
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7652379274368286,
+      "learning_rate": 4.6196404969600845e-05,
+      "loss": 1.4542,
+      "step": 14399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3794689178466797,
+      "learning_rate": 4.6196140629130326e-05,
+      "loss": 1.4431,
+      "step": 14400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8556842803955078,
+      "learning_rate": 4.619587628865979e-05,
+      "loss": 1.4314,
+      "step": 14401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.435528039932251,
+      "learning_rate": 4.619561194818927e-05,
+      "loss": 1.4401,
+      "step": 14402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.673422932624817,
+      "learning_rate": 4.619534760771874e-05,
+      "loss": 1.3756,
+      "step": 14403
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0943853855133057,
+      "learning_rate": 4.619508326724822e-05,
+      "loss": 1.3185,
+      "step": 14404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6673740148544312,
+      "learning_rate": 4.6194818926777694e-05,
+      "loss": 1.3035,
+      "step": 14405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.648409366607666,
+      "learning_rate": 4.619455458630717e-05,
+      "loss": 1.4176,
+      "step": 14406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.881648063659668,
+      "learning_rate": 4.619429024583664e-05,
+      "loss": 1.4444,
+      "step": 14407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.688913106918335,
+      "learning_rate": 4.6194025905366114e-05,
+      "loss": 1.4763,
+      "step": 14408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.873313546180725,
+      "learning_rate": 4.619376156489559e-05,
+      "loss": 1.3096,
+      "step": 14409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.12483286857605,
+      "learning_rate": 4.619349722442506e-05,
+      "loss": 1.4,
+      "step": 14410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0642776489257812,
+      "learning_rate": 4.6193232883954535e-05,
+      "loss": 1.4973,
+      "step": 14411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.578204870223999,
+      "learning_rate": 4.619296854348401e-05,
+      "loss": 1.3537,
+      "step": 14412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6063404083251953,
+      "learning_rate": 4.619270420301348e-05,
+      "loss": 1.2979,
+      "step": 14413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8241158723831177,
+      "learning_rate": 4.6192439862542956e-05,
+      "loss": 1.4451,
+      "step": 14414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.752748727798462,
+      "learning_rate": 4.6192175522072436e-05,
+      "loss": 1.5183,
+      "step": 14415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9432369470596313,
+      "learning_rate": 4.61919111816019e-05,
+      "loss": 1.4401,
+      "step": 14416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1661863327026367,
+      "learning_rate": 4.619164684113138e-05,
+      "loss": 1.4472,
+      "step": 14417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.513729214668274,
+      "learning_rate": 4.619138250066085e-05,
+      "loss": 1.367,
+      "step": 14418
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.396385431289673,
+      "learning_rate": 4.619111816019033e-05,
+      "loss": 1.4934,
+      "step": 14419
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5500701665878296,
+      "learning_rate": 4.6190853819719804e-05,
+      "loss": 1.5397,
+      "step": 14420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.361435890197754,
+      "learning_rate": 4.619058947924928e-05,
+      "loss": 1.2587,
+      "step": 14421
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9789462089538574,
+      "learning_rate": 4.619032513877875e-05,
+      "loss": 1.5768,
+      "step": 14422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.609970211982727,
+      "learning_rate": 4.619006079830822e-05,
+      "loss": 1.1917,
+      "step": 14423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6929110288619995,
+      "learning_rate": 4.61897964578377e-05,
+      "loss": 1.3567,
+      "step": 14424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9979658126831055,
+      "learning_rate": 4.618953211736717e-05,
+      "loss": 1.4934,
+      "step": 14425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.553415298461914,
+      "learning_rate": 4.6189267776896645e-05,
+      "loss": 1.523,
+      "step": 14426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.940682291984558,
+      "learning_rate": 4.618900343642612e-05,
+      "loss": 1.2452,
+      "step": 14427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5356855392456055,
+      "learning_rate": 4.618873909595559e-05,
+      "loss": 1.2731,
+      "step": 14428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8239887952804565,
+      "learning_rate": 4.6188474755485066e-05,
+      "loss": 1.5395,
+      "step": 14429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7100248336791992,
+      "learning_rate": 4.6188210415014546e-05,
+      "loss": 1.2745,
+      "step": 14430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.127962827682495,
+      "learning_rate": 4.618794607454401e-05,
+      "loss": 1.4897,
+      "step": 14431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4738335609436035,
+      "learning_rate": 4.618768173407349e-05,
+      "loss": 1.4034,
+      "step": 14432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.622829794883728,
+      "learning_rate": 4.618741739360296e-05,
+      "loss": 1.3632,
+      "step": 14433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6420060396194458,
+      "learning_rate": 4.6187153053132434e-05,
+      "loss": 1.414,
+      "step": 14434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3306711912155151,
+      "learning_rate": 4.6186888712661914e-05,
+      "loss": 1.1693,
+      "step": 14435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7284605503082275,
+      "learning_rate": 4.618662437219138e-05,
+      "loss": 1.3924,
+      "step": 14436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6226972341537476,
+      "learning_rate": 4.618636003172086e-05,
+      "loss": 1.445,
+      "step": 14437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.6572465896606445,
+      "learning_rate": 4.618609569125033e-05,
+      "loss": 1.8146,
+      "step": 14438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5631145238876343,
+      "learning_rate": 4.618583135077981e-05,
+      "loss": 1.2289,
+      "step": 14439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5427305698394775,
+      "learning_rate": 4.618556701030928e-05,
+      "loss": 1.4922,
+      "step": 14440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9778310060501099,
+      "learning_rate": 4.6185302669838756e-05,
+      "loss": 1.4531,
+      "step": 14441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5234405994415283,
+      "learning_rate": 4.618503832936823e-05,
+      "loss": 1.313,
+      "step": 14442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9683616161346436,
+      "learning_rate": 4.61847739888977e-05,
+      "loss": 1.3705,
+      "step": 14443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.19114089012146,
+      "learning_rate": 4.6184509648427176e-05,
+      "loss": 1.5411,
+      "step": 14444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4588539600372314,
+      "learning_rate": 4.618424530795665e-05,
+      "loss": 1.3276,
+      "step": 14445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5955655574798584,
+      "learning_rate": 4.6183980967486124e-05,
+      "loss": 1.4673,
+      "step": 14446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.272745132446289,
+      "learning_rate": 4.61837166270156e-05,
+      "loss": 1.6011,
+      "step": 14447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.287426710128784,
+      "learning_rate": 4.618345228654507e-05,
+      "loss": 1.4117,
+      "step": 14448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.002150535583496,
+      "learning_rate": 4.6183187946074544e-05,
+      "loss": 1.5275,
+      "step": 14449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.900655746459961,
+      "learning_rate": 4.6182923605604025e-05,
+      "loss": 1.553,
+      "step": 14450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7315006256103516,
+      "learning_rate": 4.618265926513349e-05,
+      "loss": 1.5031,
+      "step": 14451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3603901863098145,
+      "learning_rate": 4.618239492466297e-05,
+      "loss": 1.4568,
+      "step": 14452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5857436656951904,
+      "learning_rate": 4.618213058419244e-05,
+      "loss": 1.358,
+      "step": 14453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9494174718856812,
+      "learning_rate": 4.618186624372192e-05,
+      "loss": 1.4725,
+      "step": 14454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.494271993637085,
+      "learning_rate": 4.618160190325139e-05,
+      "loss": 1.5044,
+      "step": 14455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.646901249885559,
+      "learning_rate": 4.6181337562780866e-05,
+      "loss": 1.2518,
+      "step": 14456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.621633768081665,
+      "learning_rate": 4.618107322231034e-05,
+      "loss": 1.3996,
+      "step": 14457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.130542278289795,
+      "learning_rate": 4.618080888183981e-05,
+      "loss": 1.4425,
+      "step": 14458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9529309272766113,
+      "learning_rate": 4.618054454136929e-05,
+      "loss": 1.3581,
+      "step": 14459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8308393955230713,
+      "learning_rate": 4.618028020089876e-05,
+      "loss": 1.4906,
+      "step": 14460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6768393516540527,
+      "learning_rate": 4.6180015860428234e-05,
+      "loss": 1.3471,
+      "step": 14461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.330918788909912,
+      "learning_rate": 4.617975151995771e-05,
+      "loss": 1.2606,
+      "step": 14462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4984948635101318,
+      "learning_rate": 4.617948717948718e-05,
+      "loss": 1.3999,
+      "step": 14463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.35861873626709,
+      "learning_rate": 4.6179222839016655e-05,
+      "loss": 1.4,
+      "step": 14464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6304190158843994,
+      "learning_rate": 4.6178958498546135e-05,
+      "loss": 1.3401,
+      "step": 14465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4666475057601929,
+      "learning_rate": 4.61786941580756e-05,
+      "loss": 1.3918,
+      "step": 14466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.700586199760437,
+      "learning_rate": 4.617842981760508e-05,
+      "loss": 1.6618,
+      "step": 14467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5304275751113892,
+      "learning_rate": 4.617816547713455e-05,
+      "loss": 1.3792,
+      "step": 14468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.447863221168518,
+      "learning_rate": 4.617790113666402e-05,
+      "loss": 1.3291,
+      "step": 14469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9112762212753296,
+      "learning_rate": 4.61776367961935e-05,
+      "loss": 1.4279,
+      "step": 14470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6260623931884766,
+      "learning_rate": 4.617737245572297e-05,
+      "loss": 1.4218,
+      "step": 14471
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9234310388565063,
+      "learning_rate": 4.617710811525245e-05,
+      "loss": 1.2842,
+      "step": 14472
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2930766344070435,
+      "learning_rate": 4.617684377478192e-05,
+      "loss": 1.213,
+      "step": 14473
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6411783695220947,
+      "learning_rate": 4.61765794343114e-05,
+      "loss": 1.3085,
+      "step": 14474
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.443168044090271,
+      "learning_rate": 4.617631509384087e-05,
+      "loss": 1.3261,
+      "step": 14475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.679128885269165,
+      "learning_rate": 4.6176050753370344e-05,
+      "loss": 2.267,
+      "step": 14476
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7432068586349487,
+      "learning_rate": 4.617578641289982e-05,
+      "loss": 1.4458,
+      "step": 14477
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9315428733825684,
+      "learning_rate": 4.617552207242929e-05,
+      "loss": 1.5148,
+      "step": 14478
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4882333278656006,
+      "learning_rate": 4.6175257731958765e-05,
+      "loss": 1.3518,
+      "step": 14479
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5208921432495117,
+      "learning_rate": 4.617499339148824e-05,
+      "loss": 1.4115,
+      "step": 14480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4942600727081299,
+      "learning_rate": 4.617472905101771e-05,
+      "loss": 1.2531,
+      "step": 14481
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5822360515594482,
+      "learning_rate": 4.6174464710547186e-05,
+      "loss": 1.2737,
+      "step": 14482
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9244083166122437,
+      "learning_rate": 4.617420037007666e-05,
+      "loss": 1.4588,
+      "step": 14483
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7163946628570557,
+      "learning_rate": 4.617393602960613e-05,
+      "loss": 1.3671,
+      "step": 14484
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.784175395965576,
+      "learning_rate": 4.617367168913561e-05,
+      "loss": 1.4648,
+      "step": 14485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5932608842849731,
+      "learning_rate": 4.617340734866508e-05,
+      "loss": 1.319,
+      "step": 14486
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5617458820343018,
+      "learning_rate": 4.617314300819456e-05,
+      "loss": 1.4128,
+      "step": 14487
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7824565172195435,
+      "learning_rate": 4.617287866772403e-05,
+      "loss": 2.4024,
+      "step": 14488
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2033088207244873,
+      "learning_rate": 4.617261432725351e-05,
+      "loss": 2.4091,
+      "step": 14489
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5833951234817505,
+      "learning_rate": 4.617234998678298e-05,
+      "loss": 1.3702,
+      "step": 14490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.600404977798462,
+      "learning_rate": 4.6172085646312455e-05,
+      "loss": 1.5037,
+      "step": 14491
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.454924464225769,
+      "learning_rate": 4.617182130584193e-05,
+      "loss": 1.2399,
+      "step": 14492
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3645102977752686,
+      "learning_rate": 4.61715569653714e-05,
+      "loss": 1.3664,
+      "step": 14493
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5919477939605713,
+      "learning_rate": 4.6171292624900875e-05,
+      "loss": 1.3559,
+      "step": 14494
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.059307098388672,
+      "learning_rate": 4.617102828443035e-05,
+      "loss": 1.2939,
+      "step": 14495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0088412761688232,
+      "learning_rate": 4.617076394395982e-05,
+      "loss": 1.4238,
+      "step": 14496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6811929941177368,
+      "learning_rate": 4.6170499603489296e-05,
+      "loss": 1.311,
+      "step": 14497
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7081749439239502,
+      "learning_rate": 4.617023526301877e-05,
+      "loss": 1.358,
+      "step": 14498
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8827420473098755,
+      "learning_rate": 4.616997092254824e-05,
+      "loss": 1.6496,
+      "step": 14499
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.641339659690857,
+      "learning_rate": 4.6169706582077724e-05,
+      "loss": 1.2258,
+      "step": 14500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4219701290130615,
+      "learning_rate": 4.616944224160719e-05,
+      "loss": 1.5514,
+      "step": 14501
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7911680936813354,
+      "learning_rate": 4.616917790113667e-05,
+      "loss": 1.4958,
+      "step": 14502
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0971176624298096,
+      "learning_rate": 4.616891356066614e-05,
+      "loss": 1.5515,
+      "step": 14503
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6801695823669434,
+      "learning_rate": 4.616864922019561e-05,
+      "loss": 1.3365,
+      "step": 14504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5966455936431885,
+      "learning_rate": 4.616838487972509e-05,
+      "loss": 1.5035,
+      "step": 14505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4655725955963135,
+      "learning_rate": 4.616812053925456e-05,
+      "loss": 1.227,
+      "step": 14506
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7662439346313477,
+      "learning_rate": 4.616785619878404e-05,
+      "loss": 1.422,
+      "step": 14507
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5814566612243652,
+      "learning_rate": 4.6167591858313505e-05,
+      "loss": 1.4307,
+      "step": 14508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7282363176345825,
+      "learning_rate": 4.6167327517842986e-05,
+      "loss": 1.3652,
+      "step": 14509
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7174811363220215,
+      "learning_rate": 4.616706317737246e-05,
+      "loss": 1.4651,
+      "step": 14510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5601584911346436,
+      "learning_rate": 4.616679883690193e-05,
+      "loss": 1.6245,
+      "step": 14511
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5803533792495728,
+      "learning_rate": 4.6166534496431406e-05,
+      "loss": 1.2534,
+      "step": 14512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9851434230804443,
+      "learning_rate": 4.616627015596088e-05,
+      "loss": 1.4064,
+      "step": 14513
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7378019094467163,
+      "learning_rate": 4.6166005815490354e-05,
+      "loss": 1.1853,
+      "step": 14514
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8939265012741089,
+      "learning_rate": 4.616574147501983e-05,
+      "loss": 1.3141,
+      "step": 14515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.863231897354126,
+      "learning_rate": 4.61654771345493e-05,
+      "loss": 1.4631,
+      "step": 14516
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.950226068496704,
+      "learning_rate": 4.6165212794078774e-05,
+      "loss": 1.511,
+      "step": 14517
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8549937009811401,
+      "learning_rate": 4.616494845360825e-05,
+      "loss": 1.4033,
+      "step": 14518
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3432586193084717,
+      "learning_rate": 4.616468411313772e-05,
+      "loss": 1.5665,
+      "step": 14519
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7319447994232178,
+      "learning_rate": 4.61644197726672e-05,
+      "loss": 1.4557,
+      "step": 14520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6152085065841675,
+      "learning_rate": 4.616415543219667e-05,
+      "loss": 1.5203,
+      "step": 14521
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3139593601226807,
+      "learning_rate": 4.616389109172615e-05,
+      "loss": 1.5109,
+      "step": 14522
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.263312578201294,
+      "learning_rate": 4.6163626751255616e-05,
+      "loss": 1.327,
+      "step": 14523
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6971163749694824,
+      "learning_rate": 4.6163362410785096e-05,
+      "loss": 1.2904,
+      "step": 14524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5876173973083496,
+      "learning_rate": 4.616309807031457e-05,
+      "loss": 1.5494,
+      "step": 14525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7055999040603638,
+      "learning_rate": 4.616283372984404e-05,
+      "loss": 1.6431,
+      "step": 14526
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8822840452194214,
+      "learning_rate": 4.616256938937352e-05,
+      "loss": 1.3654,
+      "step": 14527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.054928779602051,
+      "learning_rate": 4.616230504890299e-05,
+      "loss": 1.4522,
+      "step": 14528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6450828313827515,
+      "learning_rate": 4.6162040708432464e-05,
+      "loss": 1.2896,
+      "step": 14529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2335009574890137,
+      "learning_rate": 4.616177636796194e-05,
+      "loss": 1.505,
+      "step": 14530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6754610538482666,
+      "learning_rate": 4.616151202749141e-05,
+      "loss": 1.3469,
+      "step": 14531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.407760500907898,
+      "learning_rate": 4.6161247687020885e-05,
+      "loss": 1.2549,
+      "step": 14532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6846247911453247,
+      "learning_rate": 4.616098334655036e-05,
+      "loss": 1.4759,
+      "step": 14533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.071312665939331,
+      "learning_rate": 4.616071900607983e-05,
+      "loss": 1.5719,
+      "step": 14534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9975881576538086,
+      "learning_rate": 4.616045466560931e-05,
+      "loss": 1.4185,
+      "step": 14535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9853017330169678,
+      "learning_rate": 4.616019032513878e-05,
+      "loss": 1.4027,
+      "step": 14536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9265111684799194,
+      "learning_rate": 4.615992598466826e-05,
+      "loss": 1.3631,
+      "step": 14537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4625451564788818,
+      "learning_rate": 4.6159661644197726e-05,
+      "loss": 1.3296,
+      "step": 14538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.518425464630127,
+      "learning_rate": 4.6159397303727206e-05,
+      "loss": 1.3981,
+      "step": 14539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4865095615386963,
+      "learning_rate": 4.615913296325668e-05,
+      "loss": 1.2685,
+      "step": 14540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.104743003845215,
+      "learning_rate": 4.615886862278615e-05,
+      "loss": 1.3454,
+      "step": 14541
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4673360586166382,
+      "learning_rate": 4.615860428231563e-05,
+      "loss": 1.3405,
+      "step": 14542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5688740015029907,
+      "learning_rate": 4.6158339941845094e-05,
+      "loss": 1.4032,
+      "step": 14543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6933646202087402,
+      "learning_rate": 4.6158075601374574e-05,
+      "loss": 1.3701,
+      "step": 14544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.056412696838379,
+      "learning_rate": 4.615781126090405e-05,
+      "loss": 1.3947,
+      "step": 14545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7830256223678589,
+      "learning_rate": 4.615754692043352e-05,
+      "loss": 1.3985,
+      "step": 14546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6144444942474365,
+      "learning_rate": 4.6157282579962995e-05,
+      "loss": 1.489,
+      "step": 14547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8611044883728027,
+      "learning_rate": 4.615701823949247e-05,
+      "loss": 1.3959,
+      "step": 14548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.096064329147339,
+      "learning_rate": 4.615675389902194e-05,
+      "loss": 1.2848,
+      "step": 14549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5574020147323608,
+      "learning_rate": 4.6156489558551416e-05,
+      "loss": 1.4035,
+      "step": 14550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.523747682571411,
+      "learning_rate": 4.615622521808089e-05,
+      "loss": 1.4834,
+      "step": 14551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5868945121765137,
+      "learning_rate": 4.615596087761036e-05,
+      "loss": 1.5977,
+      "step": 14552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3228602409362793,
+      "learning_rate": 4.6155696537139836e-05,
+      "loss": 1.6315,
+      "step": 14553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.041283369064331,
+      "learning_rate": 4.615543219666931e-05,
+      "loss": 1.5763,
+      "step": 14554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.964693546295166,
+      "learning_rate": 4.615516785619879e-05,
+      "loss": 1.48,
+      "step": 14555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.386834979057312,
+      "learning_rate": 4.615490351572826e-05,
+      "loss": 1.3019,
+      "step": 14556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5296765565872192,
+      "learning_rate": 4.615463917525774e-05,
+      "loss": 1.3484,
+      "step": 14557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7167696952819824,
+      "learning_rate": 4.6154374834787204e-05,
+      "loss": 1.4329,
+      "step": 14558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.610109567642212,
+      "learning_rate": 4.6154110494316685e-05,
+      "loss": 1.3077,
+      "step": 14559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4558448791503906,
+      "learning_rate": 4.615384615384616e-05,
+      "loss": 1.4927,
+      "step": 14560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4925932884216309,
+      "learning_rate": 4.615358181337563e-05,
+      "loss": 1.335,
+      "step": 14561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.904128074645996,
+      "learning_rate": 4.6153317472905105e-05,
+      "loss": 1.4529,
+      "step": 14562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7877882719039917,
+      "learning_rate": 4.615305313243458e-05,
+      "loss": 1.4122,
+      "step": 14563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.135438919067383,
+      "learning_rate": 4.615278879196405e-05,
+      "loss": 1.3739,
+      "step": 14564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.536857008934021,
+      "learning_rate": 4.6152524451493526e-05,
+      "loss": 1.3275,
+      "step": 14565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6612294912338257,
+      "learning_rate": 4.6152260111023e-05,
+      "loss": 1.3758,
+      "step": 14566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.077608585357666,
+      "learning_rate": 4.615199577055247e-05,
+      "loss": 1.4385,
+      "step": 14567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6979620456695557,
+      "learning_rate": 4.615173143008195e-05,
+      "loss": 1.3671,
+      "step": 14568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7115037441253662,
+      "learning_rate": 4.615146708961142e-05,
+      "loss": 1.4554,
+      "step": 14569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8437464237213135,
+      "learning_rate": 4.61512027491409e-05,
+      "loss": 1.4309,
+      "step": 14570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1860570907592773,
+      "learning_rate": 4.615093840867037e-05,
+      "loss": 1.3887,
+      "step": 14571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9818874597549438,
+      "learning_rate": 4.615067406819985e-05,
+      "loss": 1.4087,
+      "step": 14572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8423995971679688,
+      "learning_rate": 4.6150409727729315e-05,
+      "loss": 1.4395,
+      "step": 14573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7774429321289062,
+      "learning_rate": 4.6150145387258795e-05,
+      "loss": 1.4298,
+      "step": 14574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.355400800704956,
+      "learning_rate": 4.614988104678827e-05,
+      "loss": 1.2611,
+      "step": 14575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.513526439666748,
+      "learning_rate": 4.614961670631774e-05,
+      "loss": 1.3764,
+      "step": 14576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6622910499572754,
+      "learning_rate": 4.6149352365847216e-05,
+      "loss": 1.5038,
+      "step": 14577
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5126322507858276,
+      "learning_rate": 4.614908802537668e-05,
+      "loss": 1.2761,
+      "step": 14578
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.609488606452942,
+      "learning_rate": 4.614882368490616e-05,
+      "loss": 1.2082,
+      "step": 14579
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0659987926483154,
+      "learning_rate": 4.6148559344435636e-05,
+      "loss": 1.4507,
+      "step": 14580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4858877658843994,
+      "learning_rate": 4.614829500396511e-05,
+      "loss": 1.4778,
+      "step": 14581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6571719646453857,
+      "learning_rate": 4.6148030663494583e-05,
+      "loss": 1.2037,
+      "step": 14582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6768054962158203,
+      "learning_rate": 4.614776632302406e-05,
+      "loss": 1.3896,
+      "step": 14583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8409221172332764,
+      "learning_rate": 4.614750198255353e-05,
+      "loss": 1.4952,
+      "step": 14584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8762660026550293,
+      "learning_rate": 4.614723764208301e-05,
+      "loss": 1.4252,
+      "step": 14585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.777835726737976,
+      "learning_rate": 4.614697330161248e-05,
+      "loss": 1.3861,
+      "step": 14586
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.090636968612671,
+      "learning_rate": 4.614670896114195e-05,
+      "loss": 1.585,
+      "step": 14587
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6185541152954102,
+      "learning_rate": 4.6146444620671425e-05,
+      "loss": 1.3178,
+      "step": 14588
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6450248956680298,
+      "learning_rate": 4.61461802802009e-05,
+      "loss": 1.4554,
+      "step": 14589
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.115048408508301,
+      "learning_rate": 4.614591593973038e-05,
+      "loss": 1.4043,
+      "step": 14590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7739450931549072,
+      "learning_rate": 4.6145651599259846e-05,
+      "loss": 1.3176,
+      "step": 14591
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1257216930389404,
+      "learning_rate": 4.6145387258789326e-05,
+      "loss": 1.4277,
+      "step": 14592
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6502083539962769,
+      "learning_rate": 4.614512291831879e-05,
+      "loss": 1.4053,
+      "step": 14593
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8369790315628052,
+      "learning_rate": 4.614485857784827e-05,
+      "loss": 1.5462,
+      "step": 14594
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9790611267089844,
+      "learning_rate": 4.614459423737775e-05,
+      "loss": 1.5722,
+      "step": 14595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5838736295700073,
+      "learning_rate": 4.614432989690722e-05,
+      "loss": 1.3594,
+      "step": 14596
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3029489517211914,
+      "learning_rate": 4.6144065556436694e-05,
+      "loss": 1.4187,
+      "step": 14597
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.495203971862793,
+      "learning_rate": 4.614380121596617e-05,
+      "loss": 1.0557,
+      "step": 14598
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5479880571365356,
+      "learning_rate": 4.614353687549564e-05,
+      "loss": 1.3373,
+      "step": 14599
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9330353736877441,
+      "learning_rate": 4.6143272535025115e-05,
+      "loss": 1.3374,
+      "step": 14600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7731512784957886,
+      "learning_rate": 4.614300819455459e-05,
+      "loss": 1.4392,
+      "step": 14601
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9527734518051147,
+      "learning_rate": 4.614274385408406e-05,
+      "loss": 1.6749,
+      "step": 14602
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7081797122955322,
+      "learning_rate": 4.6142479513613535e-05,
+      "loss": 1.3586,
+      "step": 14603
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7797203063964844,
+      "learning_rate": 4.614221517314301e-05,
+      "loss": 1.3089,
+      "step": 14604
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9022771120071411,
+      "learning_rate": 4.614195083267249e-05,
+      "loss": 1.3516,
+      "step": 14605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4068286418914795,
+      "learning_rate": 4.6141686492201956e-05,
+      "loss": 1.4893,
+      "step": 14606
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.458425760269165,
+      "learning_rate": 4.6141422151731436e-05,
+      "loss": 1.7372,
+      "step": 14607
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.158590078353882,
+      "learning_rate": 4.61411578112609e-05,
+      "loss": 1.4235,
+      "step": 14608
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7227661609649658,
+      "learning_rate": 4.6140893470790383e-05,
+      "loss": 1.2581,
+      "step": 14609
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2158420085906982,
+      "learning_rate": 4.614062913031986e-05,
+      "loss": 1.3721,
+      "step": 14610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9622628688812256,
+      "learning_rate": 4.614036478984933e-05,
+      "loss": 1.2961,
+      "step": 14611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7761139869689941,
+      "learning_rate": 4.6140100449378804e-05,
+      "loss": 1.4213,
+      "step": 14612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3092809915542603,
+      "learning_rate": 4.613983610890827e-05,
+      "loss": 1.3139,
+      "step": 14613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.370359420776367,
+      "learning_rate": 4.613957176843775e-05,
+      "loss": 1.6012,
+      "step": 14614
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.574364423751831,
+      "learning_rate": 4.6139307427967225e-05,
+      "loss": 1.3672,
+      "step": 14615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0936901569366455,
+      "learning_rate": 4.61390430874967e-05,
+      "loss": 2.1153,
+      "step": 14616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4633350372314453,
+      "learning_rate": 4.613877874702617e-05,
+      "loss": 1.4064,
+      "step": 14617
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7291640043258667,
+      "learning_rate": 4.6138514406555646e-05,
+      "loss": 1.3949,
+      "step": 14618
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8208943605422974,
+      "learning_rate": 4.613825006608512e-05,
+      "loss": 1.2688,
+      "step": 14619
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6584311723709106,
+      "learning_rate": 4.61379857256146e-05,
+      "loss": 1.3393,
+      "step": 14620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.966119647026062,
+      "learning_rate": 4.6137721385144066e-05,
+      "loss": 1.4083,
+      "step": 14621
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.076037645339966,
+      "learning_rate": 4.613745704467354e-05,
+      "loss": 1.5545,
+      "step": 14622
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9345183372497559,
+      "learning_rate": 4.6137192704203013e-05,
+      "loss": 1.4293,
+      "step": 14623
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2167768478393555,
+      "learning_rate": 4.613692836373249e-05,
+      "loss": 1.4617,
+      "step": 14624
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6309813261032104,
+      "learning_rate": 4.613666402326197e-05,
+      "loss": 1.3456,
+      "step": 14625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0096395015716553,
+      "learning_rate": 4.6136399682791434e-05,
+      "loss": 1.4483,
+      "step": 14626
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9299107789993286,
+      "learning_rate": 4.6136135342320914e-05,
+      "loss": 1.4033,
+      "step": 14627
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7902238368988037,
+      "learning_rate": 4.613587100185038e-05,
+      "loss": 1.4717,
+      "step": 14628
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.907404899597168,
+      "learning_rate": 4.613560666137986e-05,
+      "loss": 1.8214,
+      "step": 14629
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4071276187896729,
+      "learning_rate": 4.6135342320909335e-05,
+      "loss": 1.2036,
+      "step": 14630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5018978118896484,
+      "learning_rate": 4.613507798043881e-05,
+      "loss": 1.5079,
+      "step": 14631
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.195235252380371,
+      "learning_rate": 4.613481363996828e-05,
+      "loss": 1.4697,
+      "step": 14632
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9866089820861816,
+      "learning_rate": 4.6134549299497756e-05,
+      "loss": 1.3309,
+      "step": 14633
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.372185707092285,
+      "learning_rate": 4.613428495902723e-05,
+      "loss": 1.5471,
+      "step": 14634
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0447211265563965,
+      "learning_rate": 4.61340206185567e-05,
+      "loss": 1.4762,
+      "step": 14635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.733329176902771,
+      "learning_rate": 4.613375627808618e-05,
+      "loss": 1.3353,
+      "step": 14636
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.635952949523926,
+      "learning_rate": 4.613349193761565e-05,
+      "loss": 1.504,
+      "step": 14637
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8481957912445068,
+      "learning_rate": 4.6133227597145124e-05,
+      "loss": 1.3331,
+      "step": 14638
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9326108694076538,
+      "learning_rate": 4.61329632566746e-05,
+      "loss": 1.4986,
+      "step": 14639
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4980676174163818,
+      "learning_rate": 4.613269891620408e-05,
+      "loss": 1.2771,
+      "step": 14640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6624630689620972,
+      "learning_rate": 4.6132434575733544e-05,
+      "loss": 1.3795,
+      "step": 14641
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6539386510849,
+      "learning_rate": 4.6132170235263025e-05,
+      "loss": 1.3292,
+      "step": 14642
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9699910879135132,
+      "learning_rate": 4.613190589479249e-05,
+      "loss": 1.3735,
+      "step": 14643
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.183655261993408,
+      "learning_rate": 4.613164155432197e-05,
+      "loss": 1.5449,
+      "step": 14644
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.632606863975525,
+      "learning_rate": 4.6131377213851446e-05,
+      "loss": 1.8614,
+      "step": 14645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1361963748931885,
+      "learning_rate": 4.613111287338092e-05,
+      "loss": 2.4467,
+      "step": 14646
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.990477442741394,
+      "learning_rate": 4.613084853291039e-05,
+      "loss": 1.4957,
+      "step": 14647
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.631850242614746,
+      "learning_rate": 4.6130584192439866e-05,
+      "loss": 1.5078,
+      "step": 14648
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8929849863052368,
+      "learning_rate": 4.613031985196934e-05,
+      "loss": 1.3505,
+      "step": 14649
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6011450290679932,
+      "learning_rate": 4.6130055511498813e-05,
+      "loss": 1.2546,
+      "step": 14650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3212497234344482,
+      "learning_rate": 4.612979117102829e-05,
+      "loss": 1.4505,
+      "step": 14651
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.910894513130188,
+      "learning_rate": 4.612952683055776e-05,
+      "loss": 1.3461,
+      "step": 14652
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9828288555145264,
+      "learning_rate": 4.6129262490087234e-05,
+      "loss": 1.3492,
+      "step": 14653
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8894176483154297,
+      "learning_rate": 4.612899814961671e-05,
+      "loss": 1.3011,
+      "step": 14654
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8822712898254395,
+      "learning_rate": 4.612873380914619e-05,
+      "loss": 1.5388,
+      "step": 14655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.041199207305908,
+      "learning_rate": 4.6128469468675655e-05,
+      "loss": 1.459,
+      "step": 14656
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6062933206558228,
+      "learning_rate": 4.6128205128205135e-05,
+      "loss": 1.4164,
+      "step": 14657
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8875893354415894,
+      "learning_rate": 4.61279407877346e-05,
+      "loss": 1.2701,
+      "step": 14658
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9585729837417603,
+      "learning_rate": 4.6127676447264076e-05,
+      "loss": 1.5456,
+      "step": 14659
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8119131326675415,
+      "learning_rate": 4.6127412106793556e-05,
+      "loss": 1.3417,
+      "step": 14660
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.117985248565674,
+      "learning_rate": 4.612714776632302e-05,
+      "loss": 1.4354,
+      "step": 14661
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1149895191192627,
+      "learning_rate": 4.61268834258525e-05,
+      "loss": 1.5251,
+      "step": 14662
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.162607192993164,
+      "learning_rate": 4.612661908538197e-05,
+      "loss": 1.4553,
+      "step": 14663
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5918633937835693,
+      "learning_rate": 4.612635474491145e-05,
+      "loss": 1.6225,
+      "step": 14664
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7183369398117065,
+      "learning_rate": 4.6126090404440924e-05,
+      "loss": 1.4014,
+      "step": 14665
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.816781759262085,
+      "learning_rate": 4.61258260639704e-05,
+      "loss": 1.3721,
+      "step": 14666
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1533968448638916,
+      "learning_rate": 4.612556172349987e-05,
+      "loss": 1.5181,
+      "step": 14667
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6889787912368774,
+      "learning_rate": 4.6125297383029344e-05,
+      "loss": 1.3209,
+      "step": 14668
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5703920125961304,
+      "learning_rate": 4.612503304255882e-05,
+      "loss": 1.1925,
+      "step": 14669
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8852519989013672,
+      "learning_rate": 4.612476870208829e-05,
+      "loss": 1.5209,
+      "step": 14670
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5798991918563843,
+      "learning_rate": 4.6124504361617765e-05,
+      "loss": 1.4995,
+      "step": 14671
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0159826278686523,
+      "learning_rate": 4.612424002114724e-05,
+      "loss": 1.4858,
+      "step": 14672
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.655599594116211,
+      "learning_rate": 4.612397568067671e-05,
+      "loss": 1.3295,
+      "step": 14673
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5101137161254883,
+      "learning_rate": 4.6123711340206186e-05,
+      "loss": 1.4924,
+      "step": 14674
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.963843822479248,
+      "learning_rate": 4.6123446999735666e-05,
+      "loss": 1.2845,
+      "step": 14675
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4227187633514404,
+      "learning_rate": 4.612318265926513e-05,
+      "loss": 1.3732,
+      "step": 14676
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.578444242477417,
+      "learning_rate": 4.612291831879461e-05,
+      "loss": 1.224,
+      "step": 14677
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.972005844116211,
+      "learning_rate": 4.612265397832408e-05,
+      "loss": 1.4995,
+      "step": 14678
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5611463785171509,
+      "learning_rate": 4.612238963785356e-05,
+      "loss": 1.717,
+      "step": 14679
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8825442790985107,
+      "learning_rate": 4.6122125297383034e-05,
+      "loss": 1.499,
+      "step": 14680
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.894152045249939,
+      "learning_rate": 4.612186095691251e-05,
+      "loss": 1.4154,
+      "step": 14681
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9038722515106201,
+      "learning_rate": 4.612159661644198e-05,
+      "loss": 2.1168,
+      "step": 14682
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.561773419380188,
+      "learning_rate": 4.6121332275971455e-05,
+      "loss": 1.3784,
+      "step": 14683
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.823503017425537,
+      "learning_rate": 4.612106793550093e-05,
+      "loss": 1.4825,
+      "step": 14684
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4302613735198975,
+      "learning_rate": 4.61208035950304e-05,
+      "loss": 1.3148,
+      "step": 14685
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7233132123947144,
+      "learning_rate": 4.6120539254559876e-05,
+      "loss": 1.4606,
+      "step": 14686
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8584810495376587,
+      "learning_rate": 4.612027491408935e-05,
+      "loss": 1.4426,
+      "step": 14687
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.866936206817627,
+      "learning_rate": 4.612001057361882e-05,
+      "loss": 1.3541,
+      "step": 14688
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.354332447052002,
+      "learning_rate": 4.6119746233148296e-05,
+      "loss": 1.6119,
+      "step": 14689
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8181109428405762,
+      "learning_rate": 4.6119481892677777e-05,
+      "loss": 1.5393,
+      "step": 14690
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2569150924682617,
+      "learning_rate": 4.611921755220724e-05,
+      "loss": 1.1897,
+      "step": 14691
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1802492141723633,
+      "learning_rate": 4.6118953211736724e-05,
+      "loss": 1.4821,
+      "step": 14692
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3580117225646973,
+      "learning_rate": 4.611868887126619e-05,
+      "loss": 1.4561,
+      "step": 14693
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.373420238494873,
+      "learning_rate": 4.611842453079567e-05,
+      "loss": 1.6763,
+      "step": 14694
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.102811098098755,
+      "learning_rate": 4.6118160190325144e-05,
+      "loss": 1.2928,
+      "step": 14695
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.089195489883423,
+      "learning_rate": 4.611789584985461e-05,
+      "loss": 1.3874,
+      "step": 14696
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.835968255996704,
+      "learning_rate": 4.611763150938409e-05,
+      "loss": 1.3511,
+      "step": 14697
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2270214557647705,
+      "learning_rate": 4.611736716891356e-05,
+      "loss": 1.3282,
+      "step": 14698
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5438878536224365,
+      "learning_rate": 4.611710282844304e-05,
+      "loss": 1.3771,
+      "step": 14699
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7123773097991943,
+      "learning_rate": 4.611683848797251e-05,
+      "loss": 1.283,
+      "step": 14700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6545356512069702,
+      "learning_rate": 4.6116574147501986e-05,
+      "loss": 1.264,
+      "step": 14701
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3709523677825928,
+      "learning_rate": 4.611630980703146e-05,
+      "loss": 1.4429,
+      "step": 14702
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.544735908508301,
+      "learning_rate": 4.611604546656093e-05,
+      "loss": 1.5457,
+      "step": 14703
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4318671226501465,
+      "learning_rate": 4.6115781126090407e-05,
+      "loss": 2.2181,
+      "step": 14704
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.242312431335449,
+      "learning_rate": 4.611551678561988e-05,
+      "loss": 1.3088,
+      "step": 14705
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3652045726776123,
+      "learning_rate": 4.6115252445149354e-05,
+      "loss": 1.4765,
+      "step": 14706
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7280364036560059,
+      "learning_rate": 4.611498810467883e-05,
+      "loss": 1.4252,
+      "step": 14707
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9113092422485352,
+      "learning_rate": 4.61147237642083e-05,
+      "loss": 1.3595,
+      "step": 14708
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4241167306900024,
+      "learning_rate": 4.6114459423737774e-05,
+      "loss": 1.4159,
+      "step": 14709
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8785432577133179,
+      "learning_rate": 4.6114195083267255e-05,
+      "loss": 1.3871,
+      "step": 14710
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.581381916999817,
+      "learning_rate": 4.611393074279672e-05,
+      "loss": 1.3239,
+      "step": 14711
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3040165901184082,
+      "learning_rate": 4.61136664023262e-05,
+      "loss": 1.2602,
+      "step": 14712
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6869330406188965,
+      "learning_rate": 4.611340206185567e-05,
+      "loss": 1.3629,
+      "step": 14713
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2845633029937744,
+      "learning_rate": 4.611313772138515e-05,
+      "loss": 1.5337,
+      "step": 14714
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.918560266494751,
+      "learning_rate": 4.611287338091462e-05,
+      "loss": 1.3098,
+      "step": 14715
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6866050958633423,
+      "learning_rate": 4.6112609040444096e-05,
+      "loss": 1.2371,
+      "step": 14716
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4386093616485596,
+      "learning_rate": 4.611234469997357e-05,
+      "loss": 2.1783,
+      "step": 14717
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.497938871383667,
+      "learning_rate": 4.611208035950304e-05,
+      "loss": 1.4562,
+      "step": 14718
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8242815732955933,
+      "learning_rate": 4.611181601903252e-05,
+      "loss": 1.4926,
+      "step": 14719
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9566500186920166,
+      "learning_rate": 4.611155167856199e-05,
+      "loss": 1.9185,
+      "step": 14720
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.701926827430725,
+      "learning_rate": 4.6111287338091464e-05,
+      "loss": 1.3111,
+      "step": 14721
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.058342456817627,
+      "learning_rate": 4.611102299762094e-05,
+      "loss": 1.5649,
+      "step": 14722
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.65536367893219,
+      "learning_rate": 4.611075865715041e-05,
+      "loss": 1.4207,
+      "step": 14723
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.70762038230896,
+      "learning_rate": 4.6110494316679885e-05,
+      "loss": 1.4436,
+      "step": 14724
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7544991970062256,
+      "learning_rate": 4.6110229976209365e-05,
+      "loss": 1.4299,
+      "step": 14725
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6016701459884644,
+      "learning_rate": 4.610996563573883e-05,
+      "loss": 1.3787,
+      "step": 14726
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.584836721420288,
+      "learning_rate": 4.610970129526831e-05,
+      "loss": 1.2539,
+      "step": 14727
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2576963901519775,
+      "learning_rate": 4.610943695479778e-05,
+      "loss": 1.5452,
+      "step": 14728
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8256659507751465,
+      "learning_rate": 4.610917261432726e-05,
+      "loss": 1.3878,
+      "step": 14729
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6188859939575195,
+      "learning_rate": 4.610890827385673e-05,
+      "loss": 1.5133,
+      "step": 14730
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.863223910331726,
+      "learning_rate": 4.61086439333862e-05,
+      "loss": 1.51,
+      "step": 14731
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.864311695098877,
+      "learning_rate": 4.610837959291568e-05,
+      "loss": 1.3828,
+      "step": 14732
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0059375762939453,
+      "learning_rate": 4.610811525244515e-05,
+      "loss": 1.4905,
+      "step": 14733
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.851470708847046,
+      "learning_rate": 4.610785091197463e-05,
+      "loss": 1.2915,
+      "step": 14734
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6483935117721558,
+      "learning_rate": 4.61075865715041e-05,
+      "loss": 1.3421,
+      "step": 14735
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.436483383178711,
+      "learning_rate": 4.6107322231033574e-05,
+      "loss": 1.4238,
+      "step": 14736
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8308916091918945,
+      "learning_rate": 4.610705789056305e-05,
+      "loss": 1.3643,
+      "step": 14737
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.525220274925232,
+      "learning_rate": 4.610679355009252e-05,
+      "loss": 1.1787,
+      "step": 14738
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8305022716522217,
+      "learning_rate": 4.6106529209621995e-05,
+      "loss": 1.5083,
+      "step": 14739
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.801804780960083,
+      "learning_rate": 4.610626486915147e-05,
+      "loss": 1.3537,
+      "step": 14740
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.410172462463379,
+      "learning_rate": 4.610600052868094e-05,
+      "loss": 1.5379,
+      "step": 14741
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4468110799789429,
+      "learning_rate": 4.6105736188210416e-05,
+      "loss": 1.3188,
+      "step": 14742
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.495437502861023,
+      "learning_rate": 4.610547184773989e-05,
+      "loss": 1.3087,
+      "step": 14743
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.866155982017517,
+      "learning_rate": 4.610520750726936e-05,
+      "loss": 1.5098,
+      "step": 14744
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5755691528320312,
+      "learning_rate": 4.610494316679884e-05,
+      "loss": 1.3499,
+      "step": 14745
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9436008930206299,
+      "learning_rate": 4.610467882632831e-05,
+      "loss": 1.4045,
+      "step": 14746
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6979886293411255,
+      "learning_rate": 4.610441448585779e-05,
+      "loss": 1.3436,
+      "step": 14747
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5253448486328125,
+      "learning_rate": 4.610415014538726e-05,
+      "loss": 1.4377,
+      "step": 14748
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9274121522903442,
+      "learning_rate": 4.610388580491674e-05,
+      "loss": 1.5599,
+      "step": 14749
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.490336298942566,
+      "learning_rate": 4.610362146444621e-05,
+      "loss": 1.3697,
+      "step": 14750
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3517051935195923,
+      "learning_rate": 4.6103357123975685e-05,
+      "loss": 1.2834,
+      "step": 14751
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9513025283813477,
+      "learning_rate": 4.610309278350516e-05,
+      "loss": 1.4262,
+      "step": 14752
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3589022159576416,
+      "learning_rate": 4.610282844303463e-05,
+      "loss": 1.4653,
+      "step": 14753
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7813773155212402,
+      "learning_rate": 4.6102564102564105e-05,
+      "loss": 1.2247,
+      "step": 14754
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7394354343414307,
+      "learning_rate": 4.610229976209358e-05,
+      "loss": 1.3908,
+      "step": 14755
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5036917924880981,
+      "learning_rate": 4.610203542162305e-05,
+      "loss": 1.9381,
+      "step": 14756
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.103961229324341,
+      "learning_rate": 4.6101771081152526e-05,
+      "loss": 1.5141,
+      "step": 14757
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.547935724258423,
+      "learning_rate": 4.6101506740682e-05,
+      "loss": 1.7167,
+      "step": 14758
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5512454509735107,
+      "learning_rate": 4.610124240021147e-05,
+      "loss": 1.3389,
+      "step": 14759
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6152232885360718,
+      "learning_rate": 4.6100978059740954e-05,
+      "loss": 1.265,
+      "step": 14760
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6570864915847778,
+      "learning_rate": 4.610071371927042e-05,
+      "loss": 1.4314,
+      "step": 14761
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6660610437393188,
+      "learning_rate": 4.61004493787999e-05,
+      "loss": 1.4081,
+      "step": 14762
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7111436128616333,
+      "learning_rate": 4.610018503832937e-05,
+      "loss": 1.4612,
+      "step": 14763
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4159671068191528,
+      "learning_rate": 4.609992069785885e-05,
+      "loss": 1.2246,
+      "step": 14764
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.880005121231079,
+      "learning_rate": 4.609965635738832e-05,
+      "loss": 1.4743,
+      "step": 14765
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.530418872833252,
+      "learning_rate": 4.6099392016917795e-05,
+      "loss": 1.2693,
+      "step": 14766
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7836508750915527,
+      "learning_rate": 4.609912767644727e-05,
+      "loss": 1.4758,
+      "step": 14767
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6849955320358276,
+      "learning_rate": 4.6098863335976735e-05,
+      "loss": 1.3368,
+      "step": 14768
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2796919345855713,
+      "learning_rate": 4.6098598995506216e-05,
+      "loss": 1.4022,
+      "step": 14769
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6069024801254272,
+      "learning_rate": 4.609833465503569e-05,
+      "loss": 1.2693,
+      "step": 14770
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4894025325775146,
+      "learning_rate": 4.609807031456516e-05,
+      "loss": 1.4633,
+      "step": 14771
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1314868927001953,
+      "learning_rate": 4.6097805974094636e-05,
+      "loss": 1.3826,
+      "step": 14772
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6307653188705444,
+      "learning_rate": 4.609754163362411e-05,
+      "loss": 1.3116,
+      "step": 14773
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6852425336837769,
+      "learning_rate": 4.6097277293153584e-05,
+      "loss": 1.4059,
+      "step": 14774
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9047355651855469,
+      "learning_rate": 4.6097012952683064e-05,
+      "loss": 1.3742,
+      "step": 14775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8466756343841553,
+      "learning_rate": 4.609674861221253e-05,
+      "loss": 1.4043,
+      "step": 14776
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3011233806610107,
+      "learning_rate": 4.6096484271742004e-05,
+      "loss": 1.4578,
+      "step": 14777
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.813096046447754,
+      "learning_rate": 4.609621993127148e-05,
+      "loss": 1.4698,
+      "step": 14778
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6360028982162476,
+      "learning_rate": 4.609595559080095e-05,
+      "loss": 1.4937,
+      "step": 14779
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7991586923599243,
+      "learning_rate": 4.609569125033043e-05,
+      "loss": 1.3804,
+      "step": 14780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6001895666122437,
+      "learning_rate": 4.60954269098599e-05,
+      "loss": 1.2791,
+      "step": 14781
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6655290126800537,
+      "learning_rate": 4.609516256938938e-05,
+      "loss": 1.4082,
+      "step": 14782
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.488517165184021,
+      "learning_rate": 4.6094898228918846e-05,
+      "loss": 1.3208,
+      "step": 14783
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6580204963684082,
+      "learning_rate": 4.6094633888448326e-05,
+      "loss": 1.3362,
+      "step": 14784
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.738413691520691,
+      "learning_rate": 4.60943695479778e-05,
+      "loss": 1.2569,
+      "step": 14785
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.023036479949951,
+      "learning_rate": 4.609410520750727e-05,
+      "loss": 1.5153,
+      "step": 14786
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8102341890335083,
+      "learning_rate": 4.609384086703675e-05,
+      "loss": 1.4618,
+      "step": 14787
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4135022163391113,
+      "learning_rate": 4.609357652656622e-05,
+      "loss": 1.6084,
+      "step": 14788
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.095083713531494,
+      "learning_rate": 4.6093312186095694e-05,
+      "loss": 1.4741,
+      "step": 14789
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4763009548187256,
+      "learning_rate": 4.609304784562517e-05,
+      "loss": 1.4309,
+      "step": 14790
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.66628897190094,
+      "learning_rate": 4.609278350515464e-05,
+      "loss": 1.4405,
+      "step": 14791
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8194775581359863,
+      "learning_rate": 4.6092519164684115e-05,
+      "loss": 1.4492,
+      "step": 14792
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.203925609588623,
+      "learning_rate": 4.609225482421359e-05,
+      "loss": 1.4872,
+      "step": 14793
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5436396598815918,
+      "learning_rate": 4.609199048374306e-05,
+      "loss": 1.239,
+      "step": 14794
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9579758644104004,
+      "learning_rate": 4.6091726143272535e-05,
+      "loss": 1.6028,
+      "step": 14795
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.745769739151001,
+      "learning_rate": 4.609146180280201e-05,
+      "loss": 1.4102,
+      "step": 14796
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7363343238830566,
+      "learning_rate": 4.609119746233149e-05,
+      "loss": 1.3122,
+      "step": 14797
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8868157863616943,
+      "learning_rate": 4.6090933121860956e-05,
+      "loss": 1.512,
+      "step": 14798
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0191617012023926,
+      "learning_rate": 4.6090668781390436e-05,
+      "loss": 1.565,
+      "step": 14799
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7957987785339355,
+      "learning_rate": 4.60904044409199e-05,
+      "loss": 1.3828,
+      "step": 14800
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.172769546508789,
+      "learning_rate": 4.6090140100449384e-05,
+      "loss": 1.3423,
+      "step": 14801
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8707668781280518,
+      "learning_rate": 4.608987575997886e-05,
+      "loss": 1.3992,
+      "step": 14802
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.658729910850525,
+      "learning_rate": 4.6089611419508324e-05,
+      "loss": 1.3876,
+      "step": 14803
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0389914512634277,
+      "learning_rate": 4.6089347079037804e-05,
+      "loss": 1.3919,
+      "step": 14804
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8456724882125854,
+      "learning_rate": 4.608908273856727e-05,
+      "loss": 1.3559,
+      "step": 14805
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.153350591659546,
+      "learning_rate": 4.608881839809675e-05,
+      "loss": 1.6443,
+      "step": 14806
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5819666385650635,
+      "learning_rate": 4.6088554057626225e-05,
+      "loss": 1.371,
+      "step": 14807
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8553190231323242,
+      "learning_rate": 4.60882897171557e-05,
+      "loss": 1.5216,
+      "step": 14808
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4799526929855347,
+      "learning_rate": 4.608802537668517e-05,
+      "loss": 1.4081,
+      "step": 14809
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.710410714149475,
+      "learning_rate": 4.6087761036214646e-05,
+      "loss": 1.2817,
+      "step": 14810
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4084042310714722,
+      "learning_rate": 4.608749669574412e-05,
+      "loss": 1.2512,
+      "step": 14811
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.25852108001709,
+      "learning_rate": 4.60872323552736e-05,
+      "loss": 1.663,
+      "step": 14812
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7517375946044922,
+      "learning_rate": 4.6086968014803066e-05,
+      "loss": 1.3149,
+      "step": 14813
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.668039083480835,
+      "learning_rate": 4.608670367433254e-05,
+      "loss": 1.3206,
+      "step": 14814
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.312518358230591,
+      "learning_rate": 4.6086439333862014e-05,
+      "loss": 1.3026,
+      "step": 14815
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7096534967422485,
+      "learning_rate": 4.608617499339149e-05,
+      "loss": 1.6107,
+      "step": 14816
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7868932485580444,
+      "learning_rate": 4.608591065292097e-05,
+      "loss": 1.2892,
+      "step": 14817
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6963452100753784,
+      "learning_rate": 4.6085646312450434e-05,
+      "loss": 1.9589,
+      "step": 14818
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.201205253601074,
+      "learning_rate": 4.6085381971979915e-05,
+      "loss": 1.1817,
+      "step": 14819
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7692291736602783,
+      "learning_rate": 4.608511763150938e-05,
+      "loss": 1.3715,
+      "step": 14820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.612058162689209,
+      "learning_rate": 4.608485329103886e-05,
+      "loss": 1.3921,
+      "step": 14821
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3826303482055664,
+      "learning_rate": 4.6084588950568335e-05,
+      "loss": 1.4771,
+      "step": 14822
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7290724515914917,
+      "learning_rate": 4.608432461009781e-05,
+      "loss": 1.2165,
+      "step": 14823
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.872033953666687,
+      "learning_rate": 4.608406026962728e-05,
+      "loss": 1.5819,
+      "step": 14824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2649574279785156,
+      "learning_rate": 4.6083795929156756e-05,
+      "loss": 1.4368,
+      "step": 14825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6872315406799316,
+      "learning_rate": 4.608353158868623e-05,
+      "loss": 1.2866,
+      "step": 14826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.070068836212158,
+      "learning_rate": 4.60832672482157e-05,
+      "loss": 1.2863,
+      "step": 14827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.779268741607666,
+      "learning_rate": 4.608300290774518e-05,
+      "loss": 1.453,
+      "step": 14828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2257628440856934,
+      "learning_rate": 4.608273856727465e-05,
+      "loss": 1.5305,
+      "step": 14829
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.767181634902954,
+      "learning_rate": 4.6082474226804124e-05,
+      "loss": 1.5199,
+      "step": 14830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8049802780151367,
+      "learning_rate": 4.60822098863336e-05,
+      "loss": 1.4124,
+      "step": 14831
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8603644371032715,
+      "learning_rate": 4.608194554586308e-05,
+      "loss": 1.4928,
+      "step": 14832
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2071495056152344,
+      "learning_rate": 4.6081681205392545e-05,
+      "loss": 1.2334,
+      "step": 14833
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6193629503250122,
+      "learning_rate": 4.6081416864922025e-05,
+      "loss": 1.3619,
+      "step": 14834
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.120962619781494,
+      "learning_rate": 4.608115252445149e-05,
+      "loss": 1.4241,
+      "step": 14835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6086063385009766,
+      "learning_rate": 4.608088818398097e-05,
+      "loss": 1.4233,
+      "step": 14836
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.997795581817627,
+      "learning_rate": 4.6080623843510446e-05,
+      "loss": 1.4432,
+      "step": 14837
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6489993333816528,
+      "learning_rate": 4.608035950303992e-05,
+      "loss": 1.429,
+      "step": 14838
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4941797256469727,
+      "learning_rate": 4.608009516256939e-05,
+      "loss": 1.3667,
+      "step": 14839
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7743489742279053,
+      "learning_rate": 4.607983082209886e-05,
+      "loss": 1.4846,
+      "step": 14840
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2261338233947754,
+      "learning_rate": 4.607956648162834e-05,
+      "loss": 1.5386,
+      "step": 14841
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6483360528945923,
+      "learning_rate": 4.6079302141157814e-05,
+      "loss": 1.3649,
+      "step": 14842
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8203444480895996,
+      "learning_rate": 4.607903780068729e-05,
+      "loss": 1.384,
+      "step": 14843
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5947626829147339,
+      "learning_rate": 4.607877346021676e-05,
+      "loss": 1.3305,
+      "step": 14844
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2195587158203125,
+      "learning_rate": 4.6078509119746234e-05,
+      "loss": 1.3913,
+      "step": 14845
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7892543077468872,
+      "learning_rate": 4.607824477927571e-05,
+      "loss": 1.5918,
+      "step": 14846
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8861193656921387,
+      "learning_rate": 4.607798043880519e-05,
+      "loss": 1.514,
+      "step": 14847
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.243159770965576,
+      "learning_rate": 4.6077716098334655e-05,
+      "loss": 1.6253,
+      "step": 14848
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9451760053634644,
+      "learning_rate": 4.607745175786413e-05,
+      "loss": 1.4722,
+      "step": 14849
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7589588165283203,
+      "learning_rate": 4.60771874173936e-05,
+      "loss": 1.4412,
+      "step": 14850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7982161045074463,
+      "learning_rate": 4.6076923076923076e-05,
+      "loss": 1.3864,
+      "step": 14851
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.343772530555725,
+      "learning_rate": 4.6076658736452556e-05,
+      "loss": 1.2657,
+      "step": 14852
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.466251850128174,
+      "learning_rate": 4.607639439598202e-05,
+      "loss": 1.452,
+      "step": 14853
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.600702166557312,
+      "learning_rate": 4.60761300555115e-05,
+      "loss": 1.3612,
+      "step": 14854
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.062314987182617,
+      "learning_rate": 4.607586571504097e-05,
+      "loss": 1.5529,
+      "step": 14855
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4227488040924072,
+      "learning_rate": 4.607560137457045e-05,
+      "loss": 1.3587,
+      "step": 14856
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4792851209640503,
+      "learning_rate": 4.6075337034099924e-05,
+      "loss": 1.3655,
+      "step": 14857
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3663697242736816,
+      "learning_rate": 4.60750726936294e-05,
+      "loss": 1.3707,
+      "step": 14858
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4486678838729858,
+      "learning_rate": 4.607480835315887e-05,
+      "loss": 1.3345,
+      "step": 14859
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4243075847625732,
+      "learning_rate": 4.6074544012688345e-05,
+      "loss": 1.746,
+      "step": 14860
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6548227071762085,
+      "learning_rate": 4.607427967221782e-05,
+      "loss": 1.2764,
+      "step": 14861
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4718799591064453,
+      "learning_rate": 4.607401533174729e-05,
+      "loss": 1.4112,
+      "step": 14862
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.401581883430481,
+      "learning_rate": 4.6073750991276765e-05,
+      "loss": 1.3698,
+      "step": 14863
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8967756032943726,
+      "learning_rate": 4.607348665080624e-05,
+      "loss": 1.3007,
+      "step": 14864
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7851482629776,
+      "learning_rate": 4.607322231033571e-05,
+      "loss": 1.7553,
+      "step": 14865
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9202653169631958,
+      "learning_rate": 4.6072957969865186e-05,
+      "loss": 1.5374,
+      "step": 14866
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3700768947601318,
+      "learning_rate": 4.6072693629394666e-05,
+      "loss": 1.2126,
+      "step": 14867
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.288984537124634,
+      "learning_rate": 4.607242928892413e-05,
+      "loss": 1.7647,
+      "step": 14868
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7420462369918823,
+      "learning_rate": 4.6072164948453614e-05,
+      "loss": 2.117,
+      "step": 14869
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.752144694328308,
+      "learning_rate": 4.607190060798308e-05,
+      "loss": 1.4579,
+      "step": 14870
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.809610366821289,
+      "learning_rate": 4.607163626751256e-05,
+      "loss": 1.5687,
+      "step": 14871
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6208102703094482,
+      "learning_rate": 4.6071371927042034e-05,
+      "loss": 1.2275,
+      "step": 14872
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7405610084533691,
+      "learning_rate": 4.607110758657151e-05,
+      "loss": 1.3034,
+      "step": 14873
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.855021357536316,
+      "learning_rate": 4.607084324610098e-05,
+      "loss": 1.2383,
+      "step": 14874
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2555911540985107,
+      "learning_rate": 4.607057890563045e-05,
+      "loss": 1.2494,
+      "step": 14875
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6089811325073242,
+      "learning_rate": 4.607031456515993e-05,
+      "loss": 1.3566,
+      "step": 14876
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3106756210327148,
+      "learning_rate": 4.60700502246894e-05,
+      "loss": 1.3331,
+      "step": 14877
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.814050555229187,
+      "learning_rate": 4.6069785884218876e-05,
+      "loss": 1.4727,
+      "step": 14878
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6736459732055664,
+      "learning_rate": 4.606952154374835e-05,
+      "loss": 1.3398,
+      "step": 14879
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8972163200378418,
+      "learning_rate": 4.606925720327782e-05,
+      "loss": 1.4126,
+      "step": 14880
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.365451693534851,
+      "learning_rate": 4.6068992862807296e-05,
+      "loss": 1.3245,
+      "step": 14881
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.544836163520813,
+      "learning_rate": 4.606872852233678e-05,
+      "loss": 1.365,
+      "step": 14882
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7554819583892822,
+      "learning_rate": 4.6068464181866244e-05,
+      "loss": 1.5914,
+      "step": 14883
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.954748272895813,
+      "learning_rate": 4.6068199841395724e-05,
+      "loss": 1.399,
+      "step": 14884
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6148080825805664,
+      "learning_rate": 4.606793550092519e-05,
+      "loss": 1.64,
+      "step": 14885
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.530623435974121,
+      "learning_rate": 4.6067671160454664e-05,
+      "loss": 1.2377,
+      "step": 14886
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6715078353881836,
+      "learning_rate": 4.6067406819984145e-05,
+      "loss": 1.4092,
+      "step": 14887
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5587536096572876,
+      "learning_rate": 4.606714247951361e-05,
+      "loss": 1.368,
+      "step": 14888
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9697065353393555,
+      "learning_rate": 4.606687813904309e-05,
+      "loss": 1.3313,
+      "step": 14889
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.480325222015381,
+      "learning_rate": 4.606661379857256e-05,
+      "loss": 1.4383,
+      "step": 14890
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.852817416191101,
+      "learning_rate": 4.606634945810204e-05,
+      "loss": 1.8656,
+      "step": 14891
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5705615282058716,
+      "learning_rate": 4.606608511763151e-05,
+      "loss": 1.3947,
+      "step": 14892
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6539359092712402,
+      "learning_rate": 4.6065820777160986e-05,
+      "loss": 1.2939,
+      "step": 14893
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7338762283325195,
+      "learning_rate": 4.606555643669046e-05,
+      "loss": 1.3068,
+      "step": 14894
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7003034353256226,
+      "learning_rate": 4.606529209621993e-05,
+      "loss": 1.4866,
+      "step": 14895
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9696152210235596,
+      "learning_rate": 4.606502775574941e-05,
+      "loss": 1.2603,
+      "step": 14896
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7646042108535767,
+      "learning_rate": 4.606476341527888e-05,
+      "loss": 1.1656,
+      "step": 14897
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.062105417251587,
+      "learning_rate": 4.6064499074808354e-05,
+      "loss": 2.7359,
+      "step": 14898
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8643724918365479,
+      "learning_rate": 4.606423473433783e-05,
+      "loss": 1.4023,
+      "step": 14899
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1192901134490967,
+      "learning_rate": 4.60639703938673e-05,
+      "loss": 1.3968,
+      "step": 14900
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5715116262435913,
+      "learning_rate": 4.6063706053396775e-05,
+      "loss": 1.2998,
+      "step": 14901
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8094093799591064,
+      "learning_rate": 4.6063441712926255e-05,
+      "loss": 1.3756,
+      "step": 14902
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7316663265228271,
+      "learning_rate": 4.606317737245572e-05,
+      "loss": 1.4896,
+      "step": 14903
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5839146375656128,
+      "learning_rate": 4.60629130319852e-05,
+      "loss": 1.4365,
+      "step": 14904
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7746299505233765,
+      "learning_rate": 4.606264869151467e-05,
+      "loss": 1.4173,
+      "step": 14905
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7659296989440918,
+      "learning_rate": 4.606238435104415e-05,
+      "loss": 1.2902,
+      "step": 14906
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7108796834945679,
+      "learning_rate": 4.606212001057362e-05,
+      "loss": 1.4359,
+      "step": 14907
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6291319131851196,
+      "learning_rate": 4.6061855670103096e-05,
+      "loss": 1.3877,
+      "step": 14908
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.585813045501709,
+      "learning_rate": 4.606159132963257e-05,
+      "loss": 1.4015,
+      "step": 14909
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0401809215545654,
+      "learning_rate": 4.6061326989162043e-05,
+      "loss": 1.2306,
+      "step": 14910
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.98992919921875,
+      "learning_rate": 4.606106264869152e-05,
+      "loss": 1.4962,
+      "step": 14911
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6482185125350952,
+      "learning_rate": 4.606079830822099e-05,
+      "loss": 1.3301,
+      "step": 14912
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.812084436416626,
+      "learning_rate": 4.6060533967750464e-05,
+      "loss": 1.4516,
+      "step": 14913
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6046838760375977,
+      "learning_rate": 4.606026962727994e-05,
+      "loss": 1.5627,
+      "step": 14914
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8686937093734741,
+      "learning_rate": 4.606000528680941e-05,
+      "loss": 1.3688,
+      "step": 14915
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.904308319091797,
+      "learning_rate": 4.6059740946338885e-05,
+      "loss": 1.4714,
+      "step": 14916
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.144465208053589,
+      "learning_rate": 4.6059476605868365e-05,
+      "loss": 1.4316,
+      "step": 14917
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6997411251068115,
+      "learning_rate": 4.605921226539783e-05,
+      "loss": 1.4472,
+      "step": 14918
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9064323902130127,
+      "learning_rate": 4.605894792492731e-05,
+      "loss": 1.3116,
+      "step": 14919
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.515932559967041,
+      "learning_rate": 4.605868358445678e-05,
+      "loss": 1.4006,
+      "step": 14920
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.056403875350952,
+      "learning_rate": 4.605841924398625e-05,
+      "loss": 1.5165,
+      "step": 14921
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.476428508758545,
+      "learning_rate": 4.605815490351573e-05,
+      "loss": 1.2883,
+      "step": 14922
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.879852056503296,
+      "learning_rate": 4.60578905630452e-05,
+      "loss": 1.4982,
+      "step": 14923
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.789298415184021,
+      "learning_rate": 4.605762622257468e-05,
+      "loss": 1.5334,
+      "step": 14924
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8878051042556763,
+      "learning_rate": 4.605736188210415e-05,
+      "loss": 1.4372,
+      "step": 14925
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8630757331848145,
+      "learning_rate": 4.605709754163363e-05,
+      "loss": 1.2985,
+      "step": 14926
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7502880096435547,
+      "learning_rate": 4.60568332011631e-05,
+      "loss": 1.4272,
+      "step": 14927
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8407602310180664,
+      "learning_rate": 4.6056568860692575e-05,
+      "loss": 1.3001,
+      "step": 14928
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.796473264694214,
+      "learning_rate": 4.605630452022205e-05,
+      "loss": 1.5009,
+      "step": 14929
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6054832935333252,
+      "learning_rate": 4.605604017975152e-05,
+      "loss": 1.3406,
+      "step": 14930
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.370316743850708,
+      "learning_rate": 4.6055775839280995e-05,
+      "loss": 1.5344,
+      "step": 14931
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.154160737991333,
+      "learning_rate": 4.605551149881047e-05,
+      "loss": 1.9212,
+      "step": 14932
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.39714777469635,
+      "learning_rate": 4.605524715833994e-05,
+      "loss": 1.317,
+      "step": 14933
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1296114921569824,
+      "learning_rate": 4.6054982817869416e-05,
+      "loss": 1.6309,
+      "step": 14934
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.808281660079956,
+      "learning_rate": 4.605471847739889e-05,
+      "loss": 1.3172,
+      "step": 14935
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3976976871490479,
+      "learning_rate": 4.605445413692836e-05,
+      "loss": 1.2632,
+      "step": 14936
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7010846138000488,
+      "learning_rate": 4.6054189796457843e-05,
+      "loss": 1.8088,
+      "step": 14937
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5778510570526123,
+      "learning_rate": 4.605392545598731e-05,
+      "loss": 1.282,
+      "step": 14938
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.682512044906616,
+      "learning_rate": 4.605366111551679e-05,
+      "loss": 1.4261,
+      "step": 14939
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7079225778579712,
+      "learning_rate": 4.605339677504626e-05,
+      "loss": 1.38,
+      "step": 14940
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3658292293548584,
+      "learning_rate": 4.605313243457574e-05,
+      "loss": 1.5138,
+      "step": 14941
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4662418365478516,
+      "learning_rate": 4.605286809410521e-05,
+      "loss": 1.4542,
+      "step": 14942
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.707963466644287,
+      "learning_rate": 4.6052603753634685e-05,
+      "loss": 1.2841,
+      "step": 14943
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.215587854385376,
+      "learning_rate": 4.605233941316416e-05,
+      "loss": 1.4683,
+      "step": 14944
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3693413734436035,
+      "learning_rate": 4.605207507269363e-05,
+      "loss": 1.2026,
+      "step": 14945
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7109359502792358,
+      "learning_rate": 4.6051810732223106e-05,
+      "loss": 1.1802,
+      "step": 14946
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1916635036468506,
+      "learning_rate": 4.605154639175258e-05,
+      "loss": 1.5201,
+      "step": 14947
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6531838178634644,
+      "learning_rate": 4.605128205128205e-05,
+      "loss": 1.493,
+      "step": 14948
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7831969261169434,
+      "learning_rate": 4.6051017710811526e-05,
+      "loss": 1.4888,
+      "step": 14949
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7888754606246948,
+      "learning_rate": 4.6050753370341e-05,
+      "loss": 2.4165,
+      "step": 14950
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9644290208816528,
+      "learning_rate": 4.6050489029870473e-05,
+      "loss": 1.362,
+      "step": 14951
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5279992818832397,
+      "learning_rate": 4.6050224689399954e-05,
+      "loss": 1.4655,
+      "step": 14952
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9889048337936401,
+      "learning_rate": 4.604996034892942e-05,
+      "loss": 1.4925,
+      "step": 14953
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4656314849853516,
+      "learning_rate": 4.60496960084589e-05,
+      "loss": 1.3918,
+      "step": 14954
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0463547706604004,
+      "learning_rate": 4.604943166798837e-05,
+      "loss": 1.4029,
+      "step": 14955
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4624661207199097,
+      "learning_rate": 4.604916732751785e-05,
+      "loss": 2.2627,
+      "step": 14956
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1836981773376465,
+      "learning_rate": 4.604890298704732e-05,
+      "loss": 1.5512,
+      "step": 14957
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8942328691482544,
+      "learning_rate": 4.604863864657679e-05,
+      "loss": 1.6874,
+      "step": 14958
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.529575228691101,
+      "learning_rate": 4.604837430610627e-05,
+      "loss": 1.1527,
+      "step": 14959
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4540159702301025,
+      "learning_rate": 4.6048109965635736e-05,
+      "loss": 1.3627,
+      "step": 14960
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4360885620117188,
+      "learning_rate": 4.6047845625165216e-05,
+      "loss": 1.4269,
+      "step": 14961
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.150881052017212,
+      "learning_rate": 4.604758128469469e-05,
+      "loss": 1.315,
+      "step": 14962
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8507881164550781,
+      "learning_rate": 4.604731694422416e-05,
+      "loss": 1.3795,
+      "step": 14963
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.615412950515747,
+      "learning_rate": 4.604705260375364e-05,
+      "loss": 1.3584,
+      "step": 14964
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.166092872619629,
+      "learning_rate": 4.604678826328311e-05,
+      "loss": 1.2278,
+      "step": 14965
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8311235904693604,
+      "learning_rate": 4.6046523922812584e-05,
+      "loss": 1.4803,
+      "step": 14966
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3485569953918457,
+      "learning_rate": 4.604625958234206e-05,
+      "loss": 1.4749,
+      "step": 14967
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8739466667175293,
+      "learning_rate": 4.604599524187153e-05,
+      "loss": 1.5247,
+      "step": 14968
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7028770446777344,
+      "learning_rate": 4.6045730901401005e-05,
+      "loss": 1.4357,
+      "step": 14969
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1505165100097656,
+      "learning_rate": 4.604546656093048e-05,
+      "loss": 1.3988,
+      "step": 14970
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6232810020446777,
+      "learning_rate": 4.604520222045995e-05,
+      "loss": 1.3924,
+      "step": 14971
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4869524240493774,
+      "learning_rate": 4.604493787998943e-05,
+      "loss": 2.1375,
+      "step": 14972
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9868680238723755,
+      "learning_rate": 4.60446735395189e-05,
+      "loss": 1.3355,
+      "step": 14973
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4184173345565796,
+      "learning_rate": 4.604440919904838e-05,
+      "loss": 1.333,
+      "step": 14974
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8498047590255737,
+      "learning_rate": 4.6044144858577846e-05,
+      "loss": 1.3674,
+      "step": 14975
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8538926839828491,
+      "learning_rate": 4.6043880518107326e-05,
+      "loss": 2.2418,
+      "step": 14976
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9568699598312378,
+      "learning_rate": 4.60436161776368e-05,
+      "loss": 1.3673,
+      "step": 14977
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.440898060798645,
+      "learning_rate": 4.6043351837166273e-05,
+      "loss": 1.3851,
+      "step": 14978
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4838398694992065,
+      "learning_rate": 4.604308749669575e-05,
+      "loss": 1.4747,
+      "step": 14979
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.017960548400879,
+      "learning_rate": 4.604282315622522e-05,
+      "loss": 2.2608,
+      "step": 14980
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9071400165557861,
+      "learning_rate": 4.6042558815754694e-05,
+      "loss": 1.4318,
+      "step": 14981
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3605823516845703,
+      "learning_rate": 4.604229447528417e-05,
+      "loss": 1.5197,
+      "step": 14982
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4317017793655396,
+      "learning_rate": 4.604203013481364e-05,
+      "loss": 1.1211,
+      "step": 14983
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.680611491203308,
+      "learning_rate": 4.6041765794343115e-05,
+      "loss": 1.3873,
+      "step": 14984
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6743239164352417,
+      "learning_rate": 4.604150145387259e-05,
+      "loss": 1.3737,
+      "step": 14985
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1425654888153076,
+      "learning_rate": 4.604123711340206e-05,
+      "loss": 1.8056,
+      "step": 14986
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.089489698410034,
+      "learning_rate": 4.604097277293154e-05,
+      "loss": 1.3065,
+      "step": 14987
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.921950101852417,
+      "learning_rate": 4.604070843246101e-05,
+      "loss": 1.464,
+      "step": 14988
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6706464290618896,
+      "learning_rate": 4.604044409199049e-05,
+      "loss": 1.4181,
+      "step": 14989
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2590185403823853,
+      "learning_rate": 4.6040179751519956e-05,
+      "loss": 1.2469,
+      "step": 14990
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6238685846328735,
+      "learning_rate": 4.603991541104944e-05,
+      "loss": 1.2689,
+      "step": 14991
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4359849691390991,
+      "learning_rate": 4.603965107057891e-05,
+      "loss": 1.301,
+      "step": 14992
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5721802711486816,
+      "learning_rate": 4.603938673010838e-05,
+      "loss": 1.2527,
+      "step": 14993
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7464895248413086,
+      "learning_rate": 4.603912238963786e-05,
+      "loss": 1.4095,
+      "step": 14994
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2396490573883057,
+      "learning_rate": 4.6038858049167324e-05,
+      "loss": 1.4352,
+      "step": 14995
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5843902826309204,
+      "learning_rate": 4.6038593708696804e-05,
+      "loss": 1.3141,
+      "step": 14996
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7251698970794678,
+      "learning_rate": 4.603832936822628e-05,
+      "loss": 1.3251,
+      "step": 14997
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6537718772888184,
+      "learning_rate": 4.603806502775575e-05,
+      "loss": 1.2853,
+      "step": 14998
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7088474035263062,
+      "learning_rate": 4.6037800687285225e-05,
+      "loss": 1.3525,
+      "step": 14999
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4889047145843506,
+      "learning_rate": 4.60375363468147e-05,
+      "loss": 1.3251,
+      "step": 15000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8193860054016113,
+      "learning_rate": 4.603727200634417e-05,
+      "loss": 1.3714,
+      "step": 15001
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7336639165878296,
+      "learning_rate": 4.603700766587365e-05,
+      "loss": 1.3798,
+      "step": 15002
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.242784857749939,
+      "learning_rate": 4.603674332540312e-05,
+      "loss": 1.237,
+      "step": 15003
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7394922971725464,
+      "learning_rate": 4.603647898493259e-05,
+      "loss": 1.3765,
+      "step": 15004
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.923197865486145,
+      "learning_rate": 4.6036214644462067e-05,
+      "loss": 1.4095,
+      "step": 15005
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7916117906570435,
+      "learning_rate": 4.603595030399154e-05,
+      "loss": 1.3384,
+      "step": 15006
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5145542621612549,
+      "learning_rate": 4.603568596352102e-05,
+      "loss": 1.3015,
+      "step": 15007
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.786545991897583,
+      "learning_rate": 4.603542162305049e-05,
+      "loss": 1.3666,
+      "step": 15008
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1327428817749023,
+      "learning_rate": 4.603515728257997e-05,
+      "loss": 1.2982,
+      "step": 15009
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5936915874481201,
+      "learning_rate": 4.6034892942109434e-05,
+      "loss": 1.2249,
+      "step": 15010
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.526503562927246,
+      "learning_rate": 4.6034628601638915e-05,
+      "loss": 1.4331,
+      "step": 15011
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5023233890533447,
+      "learning_rate": 4.603436426116839e-05,
+      "loss": 1.2582,
+      "step": 15012
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.525951623916626,
+      "learning_rate": 4.603409992069786e-05,
+      "loss": 1.5667,
+      "step": 15013
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9139084815979004,
+      "learning_rate": 4.6033835580227336e-05,
+      "loss": 1.3814,
+      "step": 15014
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2526121139526367,
+      "learning_rate": 4.603357123975681e-05,
+      "loss": 1.339,
+      "step": 15015
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6672999858856201,
+      "learning_rate": 4.603330689928628e-05,
+      "loss": 1.3389,
+      "step": 15016
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.630548119544983,
+      "learning_rate": 4.6033042558815756e-05,
+      "loss": 1.232,
+      "step": 15017
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5086207389831543,
+      "learning_rate": 4.603277821834523e-05,
+      "loss": 1.2519,
+      "step": 15018
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5097179412841797,
+      "learning_rate": 4.60325138778747e-05,
+      "loss": 1.5228,
+      "step": 15019
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.922328233718872,
+      "learning_rate": 4.603224953740418e-05,
+      "loss": 1.2355,
+      "step": 15020
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0474812984466553,
+      "learning_rate": 4.603198519693365e-05,
+      "loss": 1.4398,
+      "step": 15021
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4987235069274902,
+      "learning_rate": 4.603172085646313e-05,
+      "loss": 1.2987,
+      "step": 15022
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.877728819847107,
+      "learning_rate": 4.60314565159926e-05,
+      "loss": 1.5858,
+      "step": 15023
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.360769510269165,
+      "learning_rate": 4.603119217552208e-05,
+      "loss": 1.1652,
+      "step": 15024
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.026416063308716,
+      "learning_rate": 4.6030927835051545e-05,
+      "loss": 1.5651,
+      "step": 15025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1623127460479736,
+      "learning_rate": 4.6030663494581025e-05,
+      "loss": 1.5431,
+      "step": 15026
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7671546936035156,
+      "learning_rate": 4.60303991541105e-05,
+      "loss": 1.4531,
+      "step": 15027
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8612902164459229,
+      "learning_rate": 4.603013481363997e-05,
+      "loss": 1.3002,
+      "step": 15028
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.797368049621582,
+      "learning_rate": 4.6029870473169446e-05,
+      "loss": 1.6057,
+      "step": 15029
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9712077379226685,
+      "learning_rate": 4.602960613269891e-05,
+      "loss": 1.6436,
+      "step": 15030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4725916385650635,
+      "learning_rate": 4.602934179222839e-05,
+      "loss": 1.513,
+      "step": 15031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.576173186302185,
+      "learning_rate": 4.6029077451757867e-05,
+      "loss": 1.2893,
+      "step": 15032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6629891395568848,
+      "learning_rate": 4.602881311128734e-05,
+      "loss": 1.4099,
+      "step": 15033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7895115613937378,
+      "learning_rate": 4.6028548770816814e-05,
+      "loss": 1.4031,
+      "step": 15034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4705942869186401,
+      "learning_rate": 4.602828443034629e-05,
+      "loss": 1.3555,
+      "step": 15035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9908336400985718,
+      "learning_rate": 4.602802008987576e-05,
+      "loss": 1.3977,
+      "step": 15036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5404809713363647,
+      "learning_rate": 4.602775574940524e-05,
+      "loss": 1.3953,
+      "step": 15037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0201807022094727,
+      "learning_rate": 4.602749140893471e-05,
+      "loss": 1.3401,
+      "step": 15038
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8520711660385132,
+      "learning_rate": 4.602722706846418e-05,
+      "loss": 1.5392,
+      "step": 15039
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.622854471206665,
+      "learning_rate": 4.6026962727993655e-05,
+      "loss": 1.3615,
+      "step": 15040
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0031890869140625,
+      "learning_rate": 4.602669838752313e-05,
+      "loss": 1.4643,
+      "step": 15041
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5745922327041626,
+      "learning_rate": 4.602643404705261e-05,
+      "loss": 1.3656,
+      "step": 15042
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4055904150009155,
+      "learning_rate": 4.6026169706582076e-05,
+      "loss": 1.3582,
+      "step": 15043
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4620870351791382,
+      "learning_rate": 4.6025905366111556e-05,
+      "loss": 1.3496,
+      "step": 15044
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0602259635925293,
+      "learning_rate": 4.602564102564102e-05,
+      "loss": 1.5346,
+      "step": 15045
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2801473140716553,
+      "learning_rate": 4.60253766851705e-05,
+      "loss": 1.4784,
+      "step": 15046
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.545486330986023,
+      "learning_rate": 4.602511234469998e-05,
+      "loss": 1.3468,
+      "step": 15047
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6719506978988647,
+      "learning_rate": 4.602484800422945e-05,
+      "loss": 1.3943,
+      "step": 15048
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8165078163146973,
+      "learning_rate": 4.6024583663758924e-05,
+      "loss": 1.4341,
+      "step": 15049
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.247394323348999,
+      "learning_rate": 4.60243193232884e-05,
+      "loss": 1.4798,
+      "step": 15050
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1305015087127686,
+      "learning_rate": 4.602405498281787e-05,
+      "loss": 1.4631,
+      "step": 15051
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.579864740371704,
+      "learning_rate": 4.6023790642347345e-05,
+      "loss": 1.4884,
+      "step": 15052
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9076107740402222,
+      "learning_rate": 4.602352630187682e-05,
+      "loss": 1.4586,
+      "step": 15053
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.714388370513916,
+      "learning_rate": 4.602326196140629e-05,
+      "loss": 1.5633,
+      "step": 15054
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3765628337860107,
+      "learning_rate": 4.6022997620935765e-05,
+      "loss": 2.1458,
+      "step": 15055
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4634101390838623,
+      "learning_rate": 4.602273328046524e-05,
+      "loss": 1.3685,
+      "step": 15056
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.758651852607727,
+      "learning_rate": 4.602246893999472e-05,
+      "loss": 1.4321,
+      "step": 15057
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9031111001968384,
+      "learning_rate": 4.6022204599524186e-05,
+      "loss": 1.3814,
+      "step": 15058
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4283447265625,
+      "learning_rate": 4.6021940259053667e-05,
+      "loss": 1.3025,
+      "step": 15059
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.363271713256836,
+      "learning_rate": 4.602167591858313e-05,
+      "loss": 1.5459,
+      "step": 15060
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7308140993118286,
+      "learning_rate": 4.6021411578112614e-05,
+      "loss": 1.2886,
+      "step": 15061
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4338310956954956,
+      "learning_rate": 4.602114723764209e-05,
+      "loss": 1.248,
+      "step": 15062
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1865646839141846,
+      "learning_rate": 4.602088289717156e-05,
+      "loss": 1.3774,
+      "step": 15063
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4089232683181763,
+      "learning_rate": 4.6020618556701034e-05,
+      "loss": 1.2667,
+      "step": 15064
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8863463401794434,
+      "learning_rate": 4.60203542162305e-05,
+      "loss": 1.3439,
+      "step": 15065
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9550447463989258,
+      "learning_rate": 4.602008987575998e-05,
+      "loss": 1.3031,
+      "step": 15066
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.109081983566284,
+      "learning_rate": 4.6019825535289455e-05,
+      "loss": 1.3944,
+      "step": 15067
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6638821363449097,
+      "learning_rate": 4.601956119481893e-05,
+      "loss": 1.4544,
+      "step": 15068
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8245694637298584,
+      "learning_rate": 4.60192968543484e-05,
+      "loss": 1.3452,
+      "step": 15069
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3504530191421509,
+      "learning_rate": 4.6019032513877876e-05,
+      "loss": 1.3384,
+      "step": 15070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4456738233566284,
+      "learning_rate": 4.601876817340735e-05,
+      "loss": 1.3169,
+      "step": 15071
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9635999202728271,
+      "learning_rate": 4.601850383293683e-05,
+      "loss": 1.4066,
+      "step": 15072
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.619060754776001,
+      "learning_rate": 4.6018239492466297e-05,
+      "loss": 1.4001,
+      "step": 15073
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6187844276428223,
+      "learning_rate": 4.601797515199578e-05,
+      "loss": 1.3213,
+      "step": 15074
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7801716327667236,
+      "learning_rate": 4.6017710811525244e-05,
+      "loss": 1.4846,
+      "step": 15075
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4549809694290161,
+      "learning_rate": 4.601744647105472e-05,
+      "loss": 1.3192,
+      "step": 15076
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7514383792877197,
+      "learning_rate": 4.60171821305842e-05,
+      "loss": 1.3384,
+      "step": 15077
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7825909852981567,
+      "learning_rate": 4.6016917790113664e-05,
+      "loss": 1.3406,
+      "step": 15078
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7186115980148315,
+      "learning_rate": 4.6016653449643145e-05,
+      "loss": 1.2874,
+      "step": 15079
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.870245099067688,
+      "learning_rate": 4.601638910917261e-05,
+      "loss": 1.3676,
+      "step": 15080
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5576368570327759,
+      "learning_rate": 4.601612476870209e-05,
+      "loss": 1.3386,
+      "step": 15081
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1934237480163574,
+      "learning_rate": 4.6015860428231565e-05,
+      "loss": 1.4451,
+      "step": 15082
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0831525325775146,
+      "learning_rate": 4.601559608776104e-05,
+      "loss": 1.308,
+      "step": 15083
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3081297874450684,
+      "learning_rate": 4.601533174729051e-05,
+      "loss": 1.5559,
+      "step": 15084
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9476629495620728,
+      "learning_rate": 4.6015067406819986e-05,
+      "loss": 1.2973,
+      "step": 15085
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7863010168075562,
+      "learning_rate": 4.601480306634946e-05,
+      "loss": 1.3291,
+      "step": 15086
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7626458406448364,
+      "learning_rate": 4.601453872587893e-05,
+      "loss": 1.3245,
+      "step": 15087
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5281506776809692,
+      "learning_rate": 4.601427438540841e-05,
+      "loss": 2.1202,
+      "step": 15088
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.674155831336975,
+      "learning_rate": 4.601401004493788e-05,
+      "loss": 1.286,
+      "step": 15089
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3295209407806396,
+      "learning_rate": 4.6013745704467354e-05,
+      "loss": 1.5827,
+      "step": 15090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7722865343093872,
+      "learning_rate": 4.601348136399683e-05,
+      "loss": 1.3742,
+      "step": 15091
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.501909613609314,
+      "learning_rate": 4.601321702352631e-05,
+      "loss": 1.3587,
+      "step": 15092
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.791301965713501,
+      "learning_rate": 4.6012952683055775e-05,
+      "loss": 1.3545,
+      "step": 15093
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7934601306915283,
+      "learning_rate": 4.6012688342585255e-05,
+      "loss": 1.4164,
+      "step": 15094
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.857669472694397,
+      "learning_rate": 4.601242400211472e-05,
+      "loss": 1.4287,
+      "step": 15095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9593969583511353,
+      "learning_rate": 4.60121596616442e-05,
+      "loss": 1.3329,
+      "step": 15096
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8813203573226929,
+      "learning_rate": 4.6011895321173676e-05,
+      "loss": 1.5262,
+      "step": 15097
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4236788749694824,
+      "learning_rate": 4.601163098070315e-05,
+      "loss": 1.1108,
+      "step": 15098
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7165073156356812,
+      "learning_rate": 4.601136664023262e-05,
+      "loss": 1.3602,
+      "step": 15099
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6566494703292847,
+      "learning_rate": 4.6011102299762097e-05,
+      "loss": 2.6524,
+      "step": 15100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5628472566604614,
+      "learning_rate": 4.601083795929157e-05,
+      "loss": 1.3042,
+      "step": 15101
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4971060752868652,
+      "learning_rate": 4.6010573618821044e-05,
+      "loss": 1.2118,
+      "step": 15102
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3515098094940186,
+      "learning_rate": 4.601030927835052e-05,
+      "loss": 1.5958,
+      "step": 15103
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.794910192489624,
+      "learning_rate": 4.601004493787999e-05,
+      "loss": 1.4514,
+      "step": 15104
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.003185749053955,
+      "learning_rate": 4.6009780597409464e-05,
+      "loss": 1.4467,
+      "step": 15105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7752989530563354,
+      "learning_rate": 4.600951625693894e-05,
+      "loss": 2.4089,
+      "step": 15106
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8749306201934814,
+      "learning_rate": 4.600925191646842e-05,
+      "loss": 1.3809,
+      "step": 15107
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6534122228622437,
+      "learning_rate": 4.6008987575997885e-05,
+      "loss": 1.1634,
+      "step": 15108
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.739858865737915,
+      "learning_rate": 4.6008723235527365e-05,
+      "loss": 1.7591,
+      "step": 15109
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7841203212738037,
+      "learning_rate": 4.600845889505683e-05,
+      "loss": 1.5918,
+      "step": 15110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8213902711868286,
+      "learning_rate": 4.6008194554586306e-05,
+      "loss": 1.3995,
+      "step": 15111
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1122772693634033,
+      "learning_rate": 4.6007930214115786e-05,
+      "loss": 1.3503,
+      "step": 15112
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4872299432754517,
+      "learning_rate": 4.600766587364525e-05,
+      "loss": 1.4299,
+      "step": 15113
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6900172233581543,
+      "learning_rate": 4.600740153317473e-05,
+      "loss": 1.321,
+      "step": 15114
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4029617309570312,
+      "learning_rate": 4.60071371927042e-05,
+      "loss": 1.6919,
+      "step": 15115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5795000791549683,
+      "learning_rate": 4.600687285223368e-05,
+      "loss": 1.2388,
+      "step": 15116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5888150930404663,
+      "learning_rate": 4.6006608511763154e-05,
+      "loss": 1.4408,
+      "step": 15117
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.847525954246521,
+      "learning_rate": 4.600634417129263e-05,
+      "loss": 1.3867,
+      "step": 15118
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9974066019058228,
+      "learning_rate": 4.60060798308221e-05,
+      "loss": 1.4506,
+      "step": 15119
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.838104248046875,
+      "learning_rate": 4.6005815490351575e-05,
+      "loss": 1.4641,
+      "step": 15120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8527040481567383,
+      "learning_rate": 4.600555114988105e-05,
+      "loss": 1.4144,
+      "step": 15121
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7634950876235962,
+      "learning_rate": 4.600528680941052e-05,
+      "loss": 1.2439,
+      "step": 15122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1849145889282227,
+      "learning_rate": 4.6005022468939995e-05,
+      "loss": 1.6556,
+      "step": 15123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0076606273651123,
+      "learning_rate": 4.600475812846947e-05,
+      "loss": 1.5038,
+      "step": 15124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2340147495269775,
+      "learning_rate": 4.600449378799894e-05,
+      "loss": 1.3729,
+      "step": 15125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8685001134872437,
+      "learning_rate": 4.6004229447528416e-05,
+      "loss": 2.9275,
+      "step": 15126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4743058681488037,
+      "learning_rate": 4.6003965107057896e-05,
+      "loss": 1.5646,
+      "step": 15127
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.204590320587158,
+      "learning_rate": 4.600370076658736e-05,
+      "loss": 1.4819,
+      "step": 15128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.251413345336914,
+      "learning_rate": 4.6003436426116844e-05,
+      "loss": 1.5678,
+      "step": 15129
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.369979739189148,
+      "learning_rate": 4.600317208564631e-05,
+      "loss": 1.2649,
+      "step": 15130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.671222448348999,
+      "learning_rate": 4.600290774517579e-05,
+      "loss": 1.3523,
+      "step": 15131
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5758553743362427,
+      "learning_rate": 4.6002643404705264e-05,
+      "loss": 2.2369,
+      "step": 15132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2737855911254883,
+      "learning_rate": 4.600237906423474e-05,
+      "loss": 1.4806,
+      "step": 15133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.734377145767212,
+      "learning_rate": 4.600211472376421e-05,
+      "loss": 1.5539,
+      "step": 15134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9597532749176025,
+      "learning_rate": 4.6001850383293685e-05,
+      "loss": 1.3641,
+      "step": 15135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9635859727859497,
+      "learning_rate": 4.600158604282316e-05,
+      "loss": 1.3216,
+      "step": 15136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.859890341758728,
+      "learning_rate": 4.600132170235263e-05,
+      "loss": 1.4947,
+      "step": 15137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6690078973770142,
+      "learning_rate": 4.6001057361882106e-05,
+      "loss": 1.2929,
+      "step": 15138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4136931896209717,
+      "learning_rate": 4.600079302141158e-05,
+      "loss": 1.3412,
+      "step": 15139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9047614336013794,
+      "learning_rate": 4.600052868094105e-05,
+      "loss": 1.5918,
+      "step": 15140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3070905208587646,
+      "learning_rate": 4.6000264340470526e-05,
+      "loss": 1.4003,
+      "step": 15141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0352578163146973,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.4396,
+      "step": 15142
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6976531744003296,
+      "learning_rate": 4.5999735659529474e-05,
+      "loss": 1.3391,
+      "step": 15143
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8770618438720703,
+      "learning_rate": 4.5999471319058954e-05,
+      "loss": 1.4079,
+      "step": 15144
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.912832498550415,
+      "learning_rate": 4.599920697858842e-05,
+      "loss": 1.3994,
+      "step": 15145
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6257660388946533,
+      "learning_rate": 4.59989426381179e-05,
+      "loss": 1.4772,
+      "step": 15146
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6566239595413208,
+      "learning_rate": 4.5998678297647375e-05,
+      "loss": 1.3203,
+      "step": 15147
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8361179828643799,
+      "learning_rate": 4.599841395717684e-05,
+      "loss": 1.5476,
+      "step": 15148
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5258408784866333,
+      "learning_rate": 4.599814961670632e-05,
+      "loss": 1.3872,
+      "step": 15149
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9791312217712402,
+      "learning_rate": 4.599788527623579e-05,
+      "loss": 1.1714,
+      "step": 15150
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.371159076690674,
+      "learning_rate": 4.599762093576527e-05,
+      "loss": 1.5531,
+      "step": 15151
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6075224876403809,
+      "learning_rate": 4.599735659529474e-05,
+      "loss": 1.2788,
+      "step": 15152
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8422569036483765,
+      "learning_rate": 4.5997092254824216e-05,
+      "loss": 1.425,
+      "step": 15153
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.787954568862915,
+      "learning_rate": 4.599682791435369e-05,
+      "loss": 1.4502,
+      "step": 15154
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.412736415863037,
+      "learning_rate": 4.599656357388316e-05,
+      "loss": 1.2673,
+      "step": 15155
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4480408430099487,
+      "learning_rate": 4.599629923341264e-05,
+      "loss": 1.2904,
+      "step": 15156
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8116298913955688,
+      "learning_rate": 4.599603489294211e-05,
+      "loss": 1.5591,
+      "step": 15157
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5989975929260254,
+      "learning_rate": 4.5995770552471584e-05,
+      "loss": 1.4959,
+      "step": 15158
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9309154748916626,
+      "learning_rate": 4.599550621200106e-05,
+      "loss": 1.463,
+      "step": 15159
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6023170948028564,
+      "learning_rate": 4.599524187153053e-05,
+      "loss": 1.3801,
+      "step": 15160
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4202792644500732,
+      "learning_rate": 4.5994977531060005e-05,
+      "loss": 1.4938,
+      "step": 15161
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9564753770828247,
+      "learning_rate": 4.5994713190589485e-05,
+      "loss": 1.4252,
+      "step": 15162
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8992310762405396,
+      "learning_rate": 4.599444885011895e-05,
+      "loss": 1.3765,
+      "step": 15163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9775289297103882,
+      "learning_rate": 4.599418450964843e-05,
+      "loss": 1.3901,
+      "step": 15164
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5324304103851318,
+      "learning_rate": 4.59939201691779e-05,
+      "loss": 1.3489,
+      "step": 15165
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9668673276901245,
+      "learning_rate": 4.599365582870738e-05,
+      "loss": 1.4389,
+      "step": 15166
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7258572578430176,
+      "learning_rate": 4.599339148823685e-05,
+      "loss": 1.6718,
+      "step": 15167
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9294449090957642,
+      "learning_rate": 4.5993127147766326e-05,
+      "loss": 1.4473,
+      "step": 15168
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7779725790023804,
+      "learning_rate": 4.59928628072958e-05,
+      "loss": 1.8912,
+      "step": 15169
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9604724645614624,
+      "learning_rate": 4.5992598466825274e-05,
+      "loss": 1.498,
+      "step": 15170
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3755863904953003,
+      "learning_rate": 4.599233412635475e-05,
+      "loss": 1.2147,
+      "step": 15171
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5442184209823608,
+      "learning_rate": 4.599206978588422e-05,
+      "loss": 1.4213,
+      "step": 15172
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6088532209396362,
+      "learning_rate": 4.5991805445413694e-05,
+      "loss": 1.5743,
+      "step": 15173
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8119142055511475,
+      "learning_rate": 4.599154110494317e-05,
+      "loss": 1.4563,
+      "step": 15174
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7360749244689941,
+      "learning_rate": 4.599127676447264e-05,
+      "loss": 1.3413,
+      "step": 15175
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0176727771759033,
+      "learning_rate": 4.5991012424002115e-05,
+      "loss": 1.4823,
+      "step": 15176
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.136265516281128,
+      "learning_rate": 4.5990748083531595e-05,
+      "loss": 1.3337,
+      "step": 15177
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.691886305809021,
+      "learning_rate": 4.599048374306106e-05,
+      "loss": 1.3612,
+      "step": 15178
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2494399547576904,
+      "learning_rate": 4.599021940259054e-05,
+      "loss": 1.4294,
+      "step": 15179
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6561968326568604,
+      "learning_rate": 4.598995506212001e-05,
+      "loss": 1.4555,
+      "step": 15180
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9017212390899658,
+      "learning_rate": 4.598969072164949e-05,
+      "loss": 1.5226,
+      "step": 15181
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7412527799606323,
+      "learning_rate": 4.598942638117896e-05,
+      "loss": 1.2736,
+      "step": 15182
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7761244773864746,
+      "learning_rate": 4.598916204070843e-05,
+      "loss": 1.7357,
+      "step": 15183
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6638561487197876,
+      "learning_rate": 4.598889770023791e-05,
+      "loss": 1.4293,
+      "step": 15184
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3974075317382812,
+      "learning_rate": 4.598863335976738e-05,
+      "loss": 1.6661,
+      "step": 15185
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6487375497817993,
+      "learning_rate": 4.598836901929686e-05,
+      "loss": 1.4361,
+      "step": 15186
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6786538362503052,
+      "learning_rate": 4.598810467882633e-05,
+      "loss": 1.3213,
+      "step": 15187
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.723381519317627,
+      "learning_rate": 4.5987840338355805e-05,
+      "loss": 1.3896,
+      "step": 15188
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2843735218048096,
+      "learning_rate": 4.598757599788528e-05,
+      "loss": 1.4444,
+      "step": 15189
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6828489303588867,
+      "learning_rate": 4.598731165741475e-05,
+      "loss": 1.2181,
+      "step": 15190
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7252144813537598,
+      "learning_rate": 4.5987047316944225e-05,
+      "loss": 1.4364,
+      "step": 15191
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9259006977081299,
+      "learning_rate": 4.5986782976473706e-05,
+      "loss": 1.4432,
+      "step": 15192
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9201817512512207,
+      "learning_rate": 4.598651863600317e-05,
+      "loss": 1.3158,
+      "step": 15193
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.497234582901001,
+      "learning_rate": 4.5986254295532646e-05,
+      "loss": 1.5575,
+      "step": 15194
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7008044719696045,
+      "learning_rate": 4.598598995506212e-05,
+      "loss": 1.6436,
+      "step": 15195
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7873454093933105,
+      "learning_rate": 4.598572561459159e-05,
+      "loss": 1.499,
+      "step": 15196
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9408327341079712,
+      "learning_rate": 4.5985461274121074e-05,
+      "loss": 1.4042,
+      "step": 15197
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.748424768447876,
+      "learning_rate": 4.598519693365054e-05,
+      "loss": 1.3141,
+      "step": 15198
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5759503841400146,
+      "learning_rate": 4.598493259318002e-05,
+      "loss": 1.3718,
+      "step": 15199
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5343832969665527,
+      "learning_rate": 4.598466825270949e-05,
+      "loss": 1.3111,
+      "step": 15200
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6446226835250854,
+      "learning_rate": 4.598440391223897e-05,
+      "loss": 1.1608,
+      "step": 15201
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6139854192733765,
+      "learning_rate": 4.598413957176844e-05,
+      "loss": 1.2673,
+      "step": 15202
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7713208198547363,
+      "learning_rate": 4.5983875231297915e-05,
+      "loss": 1.3626,
+      "step": 15203
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0362753868103027,
+      "learning_rate": 4.598361089082739e-05,
+      "loss": 1.4882,
+      "step": 15204
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7139110565185547,
+      "learning_rate": 4.598334655035686e-05,
+      "loss": 1.3519,
+      "step": 15205
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5006121397018433,
+      "learning_rate": 4.5983082209886336e-05,
+      "loss": 1.3048,
+      "step": 15206
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9442692995071411,
+      "learning_rate": 4.598281786941581e-05,
+      "loss": 1.4429,
+      "step": 15207
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1047186851501465,
+      "learning_rate": 4.598255352894528e-05,
+      "loss": 1.4367,
+      "step": 15208
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0480589866638184,
+      "learning_rate": 4.5982289188474756e-05,
+      "loss": 1.5097,
+      "step": 15209
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9380290508270264,
+      "learning_rate": 4.598202484800423e-05,
+      "loss": 1.3994,
+      "step": 15210
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5258300304412842,
+      "learning_rate": 4.5981760507533704e-05,
+      "loss": 1.2291,
+      "step": 15211
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5882632732391357,
+      "learning_rate": 4.5981496167063184e-05,
+      "loss": 1.0986,
+      "step": 15212
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9293307065963745,
+      "learning_rate": 4.598123182659265e-05,
+      "loss": 1.447,
+      "step": 15213
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8909684419631958,
+      "learning_rate": 4.598096748612213e-05,
+      "loss": 1.4106,
+      "step": 15214
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8493086099624634,
+      "learning_rate": 4.59807031456516e-05,
+      "loss": 2.1211,
+      "step": 15215
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.942674160003662,
+      "learning_rate": 4.598043880518108e-05,
+      "loss": 1.3343,
+      "step": 15216
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.968965768814087,
+      "learning_rate": 4.598017446471055e-05,
+      "loss": 1.469,
+      "step": 15217
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8321797847747803,
+      "learning_rate": 4.5979910124240025e-05,
+      "loss": 1.4699,
+      "step": 15218
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7479580640792847,
+      "learning_rate": 4.59796457837695e-05,
+      "loss": 1.3949,
+      "step": 15219
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.069195508956909,
+      "learning_rate": 4.5979381443298966e-05,
+      "loss": 1.3559,
+      "step": 15220
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6127667427062988,
+      "learning_rate": 4.5979117102828446e-05,
+      "loss": 1.406,
+      "step": 15221
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7448021173477173,
+      "learning_rate": 4.597885276235792e-05,
+      "loss": 1.3732,
+      "step": 15222
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0201575756073,
+      "learning_rate": 4.597858842188739e-05,
+      "loss": 1.4452,
+      "step": 15223
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2181293964385986,
+      "learning_rate": 4.597832408141687e-05,
+      "loss": 1.325,
+      "step": 15224
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8421131372451782,
+      "learning_rate": 4.597805974094634e-05,
+      "loss": 1.4009,
+      "step": 15225
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4737493991851807,
+      "learning_rate": 4.5977795400475814e-05,
+      "loss": 1.2518,
+      "step": 15226
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.72149658203125,
+      "learning_rate": 4.5977531060005294e-05,
+      "loss": 1.4111,
+      "step": 15227
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3454993963241577,
+      "learning_rate": 4.597726671953476e-05,
+      "loss": 1.155,
+      "step": 15228
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5183510780334473,
+      "learning_rate": 4.5977002379064235e-05,
+      "loss": 1.5787,
+      "step": 15229
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0278818607330322,
+      "learning_rate": 4.597673803859371e-05,
+      "loss": 1.4613,
+      "step": 15230
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6128590106964111,
+      "learning_rate": 4.597647369812318e-05,
+      "loss": 1.3168,
+      "step": 15231
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4371986389160156,
+      "learning_rate": 4.597620935765266e-05,
+      "loss": 1.5253,
+      "step": 15232
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0442211627960205,
+      "learning_rate": 4.597594501718213e-05,
+      "loss": 1.5307,
+      "step": 15233
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6383213996887207,
+      "learning_rate": 4.597568067671161e-05,
+      "loss": 1.7855,
+      "step": 15234
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4817392826080322,
+      "learning_rate": 4.5975416336241076e-05,
+      "loss": 1.3023,
+      "step": 15235
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.792974591255188,
+      "learning_rate": 4.5975151995770556e-05,
+      "loss": 1.4998,
+      "step": 15236
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.70412015914917,
+      "learning_rate": 4.597488765530003e-05,
+      "loss": 1.3607,
+      "step": 15237
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.536244511604309,
+      "learning_rate": 4.5974623314829504e-05,
+      "loss": 1.3477,
+      "step": 15238
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4581539630889893,
+      "learning_rate": 4.597435897435898e-05,
+      "loss": 1.3169,
+      "step": 15239
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.936784267425537,
+      "learning_rate": 4.597409463388845e-05,
+      "loss": 1.3074,
+      "step": 15240
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2510006427764893,
+      "learning_rate": 4.5973830293417924e-05,
+      "loss": 1.5596,
+      "step": 15241
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0320866107940674,
+      "learning_rate": 4.59735659529474e-05,
+      "loss": 1.4017,
+      "step": 15242
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6987240314483643,
+      "learning_rate": 4.597330161247687e-05,
+      "loss": 1.4825,
+      "step": 15243
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6481000185012817,
+      "learning_rate": 4.5973037272006345e-05,
+      "loss": 1.399,
+      "step": 15244
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5550216436386108,
+      "learning_rate": 4.597277293153582e-05,
+      "loss": 2.1265,
+      "step": 15245
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.104621410369873,
+      "learning_rate": 4.597250859106529e-05,
+      "loss": 1.5362,
+      "step": 15246
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5546703338623047,
+      "learning_rate": 4.597224425059477e-05,
+      "loss": 1.3739,
+      "step": 15247
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8141549825668335,
+      "learning_rate": 4.597197991012424e-05,
+      "loss": 1.2084,
+      "step": 15248
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5452078580856323,
+      "learning_rate": 4.597171556965372e-05,
+      "loss": 1.2306,
+      "step": 15249
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9516960382461548,
+      "learning_rate": 4.5971451229183186e-05,
+      "loss": 1.5613,
+      "step": 15250
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.172208786010742,
+      "learning_rate": 4.597118688871267e-05,
+      "loss": 1.3562,
+      "step": 15251
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5455200672149658,
+      "learning_rate": 4.597092254824214e-05,
+      "loss": 1.1997,
+      "step": 15252
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9236642122268677,
+      "learning_rate": 4.5970658207771614e-05,
+      "loss": 1.4177,
+      "step": 15253
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.087268829345703,
+      "learning_rate": 4.597039386730109e-05,
+      "loss": 1.5314,
+      "step": 15254
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.033546209335327,
+      "learning_rate": 4.597012952683056e-05,
+      "loss": 1.4565,
+      "step": 15255
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0781309604644775,
+      "learning_rate": 4.5969865186360035e-05,
+      "loss": 1.4697,
+      "step": 15256
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7936989068984985,
+      "learning_rate": 4.596960084588951e-05,
+      "loss": 1.426,
+      "step": 15257
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3927596807479858,
+      "learning_rate": 4.596933650541898e-05,
+      "loss": 1.1987,
+      "step": 15258
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.417644500732422,
+      "learning_rate": 4.5969072164948455e-05,
+      "loss": 1.4404,
+      "step": 15259
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4898273944854736,
+      "learning_rate": 4.596880782447793e-05,
+      "loss": 1.381,
+      "step": 15260
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9211997985839844,
+      "learning_rate": 4.59685434840074e-05,
+      "loss": 2.6703,
+      "step": 15261
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4399681091308594,
+      "learning_rate": 4.596827914353688e-05,
+      "loss": 1.3295,
+      "step": 15262
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5686990022659302,
+      "learning_rate": 4.596801480306635e-05,
+      "loss": 1.3491,
+      "step": 15263
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7681723833084106,
+      "learning_rate": 4.596775046259583e-05,
+      "loss": 1.4411,
+      "step": 15264
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.317758560180664,
+      "learning_rate": 4.59674861221253e-05,
+      "loss": 1.5055,
+      "step": 15265
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.287644624710083,
+      "learning_rate": 4.596722178165477e-05,
+      "loss": 1.5415,
+      "step": 15266
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7581636905670166,
+      "learning_rate": 4.596695744118425e-05,
+      "loss": 1.2138,
+      "step": 15267
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2034668922424316,
+      "learning_rate": 4.596669310071372e-05,
+      "loss": 1.373,
+      "step": 15268
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7680659294128418,
+      "learning_rate": 4.59664287602432e-05,
+      "loss": 1.1246,
+      "step": 15269
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7451287508010864,
+      "learning_rate": 4.5966164419772665e-05,
+      "loss": 1.4186,
+      "step": 15270
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.883749008178711,
+      "learning_rate": 4.5965900079302145e-05,
+      "loss": 1.4175,
+      "step": 15271
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6620923280715942,
+      "learning_rate": 4.596563573883162e-05,
+      "loss": 1.4958,
+      "step": 15272
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.792006254196167,
+      "learning_rate": 4.596537139836109e-05,
+      "loss": 1.4103,
+      "step": 15273
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6880351305007935,
+      "learning_rate": 4.5965107057890566e-05,
+      "loss": 1.3883,
+      "step": 15274
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.717424750328064,
+      "learning_rate": 4.596484271742004e-05,
+      "loss": 1.307,
+      "step": 15275
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4817843437194824,
+      "learning_rate": 4.596457837694951e-05,
+      "loss": 1.3141,
+      "step": 15276
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.676395058631897,
+      "learning_rate": 4.5964314036478986e-05,
+      "loss": 1.403,
+      "step": 15277
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5217649936676025,
+      "learning_rate": 4.596404969600846e-05,
+      "loss": 1.3432,
+      "step": 15278
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.672229528427124,
+      "learning_rate": 4.5963785355537933e-05,
+      "loss": 1.6718,
+      "step": 15279
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7833638191223145,
+      "learning_rate": 4.596352101506741e-05,
+      "loss": 1.2617,
+      "step": 15280
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8643239736557007,
+      "learning_rate": 4.596325667459688e-05,
+      "loss": 1.4212,
+      "step": 15281
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4265635013580322,
+      "learning_rate": 4.596299233412636e-05,
+      "loss": 1.3203,
+      "step": 15282
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4953100681304932,
+      "learning_rate": 4.596272799365583e-05,
+      "loss": 1.3328,
+      "step": 15283
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7322077751159668,
+      "learning_rate": 4.596246365318531e-05,
+      "loss": 1.3708,
+      "step": 15284
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.353799819946289,
+      "learning_rate": 4.5962199312714775e-05,
+      "loss": 1.1677,
+      "step": 15285
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.108104705810547,
+      "learning_rate": 4.5961934972244255e-05,
+      "loss": 1.2673,
+      "step": 15286
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.631985068321228,
+      "learning_rate": 4.596167063177373e-05,
+      "loss": 2.0262,
+      "step": 15287
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3970410823822021,
+      "learning_rate": 4.59614062913032e-05,
+      "loss": 1.1683,
+      "step": 15288
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4572649002075195,
+      "learning_rate": 4.5961141950832676e-05,
+      "loss": 1.5618,
+      "step": 15289
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9909781217575073,
+      "learning_rate": 4.596087761036215e-05,
+      "loss": 1.4684,
+      "step": 15290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2201900482177734,
+      "learning_rate": 4.596061326989162e-05,
+      "loss": 1.3556,
+      "step": 15291
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5542879104614258,
+      "learning_rate": 4.59603489294211e-05,
+      "loss": 1.3926,
+      "step": 15292
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.520779609680176,
+      "learning_rate": 4.596008458895057e-05,
+      "loss": 1.5457,
+      "step": 15293
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9215084314346313,
+      "learning_rate": 4.5959820248480044e-05,
+      "loss": 1.4572,
+      "step": 15294
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.049546003341675,
+      "learning_rate": 4.595955590800952e-05,
+      "loss": 1.3598,
+      "step": 15295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.843906283378601,
+      "learning_rate": 4.595929156753899e-05,
+      "loss": 1.434,
+      "step": 15296
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.977325201034546,
+      "learning_rate": 4.595902722706847e-05,
+      "loss": 1.3862,
+      "step": 15297
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3819527626037598,
+      "learning_rate": 4.595876288659794e-05,
+      "loss": 1.4631,
+      "step": 15298
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1823246479034424,
+      "learning_rate": 4.595849854612742e-05,
+      "loss": 1.5103,
+      "step": 15299
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7600533962249756,
+      "learning_rate": 4.5958234205656885e-05,
+      "loss": 1.3614,
+      "step": 15300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8345630168914795,
+      "learning_rate": 4.595796986518636e-05,
+      "loss": 1.4325,
+      "step": 15301
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0583198070526123,
+      "learning_rate": 4.595770552471584e-05,
+      "loss": 1.4479,
+      "step": 15302
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8201180696487427,
+      "learning_rate": 4.5957441184245306e-05,
+      "loss": 1.4519,
+      "step": 15303
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5495648384094238,
+      "learning_rate": 4.5957176843774786e-05,
+      "loss": 1.367,
+      "step": 15304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.936867356300354,
+      "learning_rate": 4.595691250330425e-05,
+      "loss": 1.3793,
+      "step": 15305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3535385131835938,
+      "learning_rate": 4.5956648162833733e-05,
+      "loss": 1.4808,
+      "step": 15306
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.599840760231018,
+      "learning_rate": 4.595638382236321e-05,
+      "loss": 2.3153,
+      "step": 15307
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1508491039276123,
+      "learning_rate": 4.595611948189268e-05,
+      "loss": 1.4634,
+      "step": 15308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5302485227584839,
+      "learning_rate": 4.5955855141422154e-05,
+      "loss": 1.189,
+      "step": 15309
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0547139644622803,
+      "learning_rate": 4.595559080095163e-05,
+      "loss": 1.5607,
+      "step": 15310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0915136337280273,
+      "learning_rate": 4.59553264604811e-05,
+      "loss": 1.3631,
+      "step": 15311
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3356484174728394,
+      "learning_rate": 4.5955062120010575e-05,
+      "loss": 1.2009,
+      "step": 15312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2786858081817627,
+      "learning_rate": 4.595479777954005e-05,
+      "loss": 1.4006,
+      "step": 15313
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8606178760528564,
+      "learning_rate": 4.595453343906952e-05,
+      "loss": 1.3315,
+      "step": 15314
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7451094388961792,
+      "learning_rate": 4.5954269098598996e-05,
+      "loss": 1.262,
+      "step": 15315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9829503297805786,
+      "learning_rate": 4.595400475812847e-05,
+      "loss": 1.4626,
+      "step": 15316
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7333948612213135,
+      "learning_rate": 4.595374041765795e-05,
+      "loss": 1.6864,
+      "step": 15317
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9838148355484009,
+      "learning_rate": 4.5953476077187416e-05,
+      "loss": 1.4012,
+      "step": 15318
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5785976648330688,
+      "learning_rate": 4.59532117367169e-05,
+      "loss": 1.305,
+      "step": 15319
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0515329837799072,
+      "learning_rate": 4.5952947396246363e-05,
+      "loss": 1.3744,
+      "step": 15320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.152491331100464,
+      "learning_rate": 4.5952683055775844e-05,
+      "loss": 1.5111,
+      "step": 15321
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.366731882095337,
+      "learning_rate": 4.595241871530532e-05,
+      "loss": 1.4467,
+      "step": 15322
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2477331161499023,
+      "learning_rate": 4.595215437483479e-05,
+      "loss": 2.754,
+      "step": 15323
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8003040552139282,
+      "learning_rate": 4.5951890034364264e-05,
+      "loss": 1.352,
+      "step": 15324
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.182263135910034,
+      "learning_rate": 4.595162569389374e-05,
+      "loss": 1.3452,
+      "step": 15325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8927514553070068,
+      "learning_rate": 4.595136135342321e-05,
+      "loss": 1.3698,
+      "step": 15326
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9166295528411865,
+      "learning_rate": 4.5951097012952685e-05,
+      "loss": 1.4672,
+      "step": 15327
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.036593437194824,
+      "learning_rate": 4.595083267248216e-05,
+      "loss": 1.2924,
+      "step": 15328
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.033555269241333,
+      "learning_rate": 4.595056833201163e-05,
+      "loss": 1.9826,
+      "step": 15329
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7563108205795288,
+      "learning_rate": 4.5950303991541106e-05,
+      "loss": 1.3296,
+      "step": 15330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8508137464523315,
+      "learning_rate": 4.595003965107058e-05,
+      "loss": 1.4766,
+      "step": 15331
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3602569103240967,
+      "learning_rate": 4.594977531060006e-05,
+      "loss": 1.5174,
+      "step": 15332
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.378714084625244,
+      "learning_rate": 4.594951097012953e-05,
+      "loss": 1.4294,
+      "step": 15333
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4854555130004883,
+      "learning_rate": 4.594924662965901e-05,
+      "loss": 1.467,
+      "step": 15334
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.745910882949829,
+      "learning_rate": 4.5948982289188474e-05,
+      "loss": 1.4606,
+      "step": 15335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9317359924316406,
+      "learning_rate": 4.5948717948717954e-05,
+      "loss": 1.5096,
+      "step": 15336
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6714122295379639,
+      "learning_rate": 4.594845360824743e-05,
+      "loss": 1.5147,
+      "step": 15337
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5781922340393066,
+      "learning_rate": 4.5948189267776894e-05,
+      "loss": 1.3338,
+      "step": 15338
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7497613430023193,
+      "learning_rate": 4.5947924927306375e-05,
+      "loss": 1.411,
+      "step": 15339
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9059381484985352,
+      "learning_rate": 4.594766058683584e-05,
+      "loss": 1.3261,
+      "step": 15340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0328545570373535,
+      "learning_rate": 4.594739624636532e-05,
+      "loss": 1.4253,
+      "step": 15341
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.499232292175293,
+      "learning_rate": 4.5947131905894796e-05,
+      "loss": 1.3604,
+      "step": 15342
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.869344711303711,
+      "learning_rate": 4.594686756542427e-05,
+      "loss": 1.564,
+      "step": 15343
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4889498949050903,
+      "learning_rate": 4.594660322495374e-05,
+      "loss": 1.3132,
+      "step": 15344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2061526775360107,
+      "learning_rate": 4.5946338884483216e-05,
+      "loss": 1.4811,
+      "step": 15345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3939621448516846,
+      "learning_rate": 4.594607454401269e-05,
+      "loss": 1.1786,
+      "step": 15346
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7697352170944214,
+      "learning_rate": 4.5945810203542163e-05,
+      "loss": 1.305,
+      "step": 15347
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9043289422988892,
+      "learning_rate": 4.594554586307164e-05,
+      "loss": 1.4676,
+      "step": 15348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9863945245742798,
+      "learning_rate": 4.594528152260111e-05,
+      "loss": 1.5163,
+      "step": 15349
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7082788944244385,
+      "learning_rate": 4.5945017182130584e-05,
+      "loss": 1.4624,
+      "step": 15350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.012723922729492,
+      "learning_rate": 4.594475284166006e-05,
+      "loss": 1.3819,
+      "step": 15351
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.069070339202881,
+      "learning_rate": 4.594448850118954e-05,
+      "loss": 1.4928,
+      "step": 15352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6549614667892456,
+      "learning_rate": 4.5944224160719005e-05,
+      "loss": 2.1303,
+      "step": 15353
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.503800392150879,
+      "learning_rate": 4.5943959820248485e-05,
+      "loss": 1.3867,
+      "step": 15354
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7483198642730713,
+      "learning_rate": 4.594369547977795e-05,
+      "loss": 1.3028,
+      "step": 15355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6859240531921387,
+      "learning_rate": 4.594343113930743e-05,
+      "loss": 1.5352,
+      "step": 15356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9988921880722046,
+      "learning_rate": 4.5943166798836906e-05,
+      "loss": 1.519,
+      "step": 15357
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2062206268310547,
+      "learning_rate": 4.594290245836638e-05,
+      "loss": 1.4019,
+      "step": 15358
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4423904418945312,
+      "learning_rate": 4.594263811789585e-05,
+      "loss": 1.283,
+      "step": 15359
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.007441997528076,
+      "learning_rate": 4.5942373777425327e-05,
+      "loss": 1.3765,
+      "step": 15360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7460849285125732,
+      "learning_rate": 4.59421094369548e-05,
+      "loss": 1.9072,
+      "step": 15361
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8028644323349,
+      "learning_rate": 4.5941845096484274e-05,
+      "loss": 1.4707,
+      "step": 15362
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2157275676727295,
+      "learning_rate": 4.594158075601375e-05,
+      "loss": 1.883,
+      "step": 15363
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1000115871429443,
+      "learning_rate": 4.594131641554322e-05,
+      "loss": 1.5733,
+      "step": 15364
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9444918632507324,
+      "learning_rate": 4.5941052075072694e-05,
+      "loss": 1.4998,
+      "step": 15365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7999765872955322,
+      "learning_rate": 4.594078773460217e-05,
+      "loss": 1.494,
+      "step": 15366
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7678364515304565,
+      "learning_rate": 4.594052339413165e-05,
+      "loss": 1.2232,
+      "step": 15367
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6070996522903442,
+      "learning_rate": 4.5940259053661115e-05,
+      "loss": 1.3692,
+      "step": 15368
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5585217475891113,
+      "learning_rate": 4.5939994713190596e-05,
+      "loss": 1.4744,
+      "step": 15369
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6559914350509644,
+      "learning_rate": 4.593973037272006e-05,
+      "loss": 1.437,
+      "step": 15370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.459718942642212,
+      "learning_rate": 4.593946603224954e-05,
+      "loss": 1.35,
+      "step": 15371
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7208172082901,
+      "learning_rate": 4.5939201691779016e-05,
+      "loss": 1.3875,
+      "step": 15372
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7351957559585571,
+      "learning_rate": 4.593893735130849e-05,
+      "loss": 1.4463,
+      "step": 15373
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5107651948928833,
+      "learning_rate": 4.593867301083796e-05,
+      "loss": 2.1563,
+      "step": 15374
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.442224144935608,
+      "learning_rate": 4.593840867036743e-05,
+      "loss": 1.299,
+      "step": 15375
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7414374351501465,
+      "learning_rate": 4.593814432989691e-05,
+      "loss": 1.4886,
+      "step": 15376
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6681406497955322,
+      "learning_rate": 4.5937879989426384e-05,
+      "loss": 1.4938,
+      "step": 15377
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.639309048652649,
+      "learning_rate": 4.593761564895586e-05,
+      "loss": 1.3524,
+      "step": 15378
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4604883193969727,
+      "learning_rate": 4.593735130848533e-05,
+      "loss": 1.2886,
+      "step": 15379
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4120749235153198,
+      "learning_rate": 4.5937086968014805e-05,
+      "loss": 1.4058,
+      "step": 15380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.947422981262207,
+      "learning_rate": 4.593682262754428e-05,
+      "loss": 1.3704,
+      "step": 15381
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.827517032623291,
+      "learning_rate": 4.593655828707376e-05,
+      "loss": 1.328,
+      "step": 15382
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6112613677978516,
+      "learning_rate": 4.5936293946603225e-05,
+      "loss": 1.418,
+      "step": 15383
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.851339340209961,
+      "learning_rate": 4.59360296061327e-05,
+      "loss": 1.4142,
+      "step": 15384
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.451539993286133,
+      "learning_rate": 4.593576526566217e-05,
+      "loss": 1.4332,
+      "step": 15385
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5965025424957275,
+      "learning_rate": 4.5935500925191646e-05,
+      "loss": 1.3981,
+      "step": 15386
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6664851903915405,
+      "learning_rate": 4.5935236584721127e-05,
+      "loss": 1.3352,
+      "step": 15387
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5335391759872437,
+      "learning_rate": 4.593497224425059e-05,
+      "loss": 1.4442,
+      "step": 15388
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0209569931030273,
+      "learning_rate": 4.5934707903780074e-05,
+      "loss": 1.2999,
+      "step": 15389
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7020691633224487,
+      "learning_rate": 4.593444356330954e-05,
+      "loss": 1.21,
+      "step": 15390
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.075145721435547,
+      "learning_rate": 4.593417922283902e-05,
+      "loss": 1.4218,
+      "step": 15391
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5487701892852783,
+      "learning_rate": 4.5933914882368494e-05,
+      "loss": 1.3603,
+      "step": 15392
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.786368489265442,
+      "learning_rate": 4.593365054189797e-05,
+      "loss": 1.4649,
+      "step": 15393
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4082889556884766,
+      "learning_rate": 4.593338620142744e-05,
+      "loss": 1.4199,
+      "step": 15394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5402451753616333,
+      "learning_rate": 4.5933121860956915e-05,
+      "loss": 1.3285,
+      "step": 15395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6292915344238281,
+      "learning_rate": 4.593285752048639e-05,
+      "loss": 1.309,
+      "step": 15396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8112443685531616,
+      "learning_rate": 4.593259318001586e-05,
+      "loss": 1.3389,
+      "step": 15397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.639201045036316,
+      "learning_rate": 4.5932328839545336e-05,
+      "loss": 1.2018,
+      "step": 15398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3354908227920532,
+      "learning_rate": 4.593206449907481e-05,
+      "loss": 1.2525,
+      "step": 15399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9037444591522217,
+      "learning_rate": 4.593180015860428e-05,
+      "loss": 1.4144,
+      "step": 15400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.060098648071289,
+      "learning_rate": 4.5931535818133757e-05,
+      "loss": 1.2745,
+      "step": 15401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2124054431915283,
+      "learning_rate": 4.593127147766324e-05,
+      "loss": 1.5414,
+      "step": 15402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.762690782546997,
+      "learning_rate": 4.5931007137192704e-05,
+      "loss": 1.3677,
+      "step": 15403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9044840335845947,
+      "learning_rate": 4.5930742796722184e-05,
+      "loss": 1.4432,
+      "step": 15404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3538074493408203,
+      "learning_rate": 4.593047845625165e-05,
+      "loss": 1.3939,
+      "step": 15405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7357354164123535,
+      "learning_rate": 4.593021411578113e-05,
+      "loss": 1.3799,
+      "step": 15406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.730292797088623,
+      "learning_rate": 4.5929949775310605e-05,
+      "loss": 1.5899,
+      "step": 15407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.966081976890564,
+      "learning_rate": 4.592968543484008e-05,
+      "loss": 1.5339,
+      "step": 15408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.811158299446106,
+      "learning_rate": 4.592942109436955e-05,
+      "loss": 1.2299,
+      "step": 15409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8141872882843018,
+      "learning_rate": 4.592915675389902e-05,
+      "loss": 1.406,
+      "step": 15410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7168735265731812,
+      "learning_rate": 4.59288924134285e-05,
+      "loss": 1.4081,
+      "step": 15411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7621513605117798,
+      "learning_rate": 4.592862807295797e-05,
+      "loss": 1.25,
+      "step": 15412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7370216846466064,
+      "learning_rate": 4.5928363732487446e-05,
+      "loss": 1.4222,
+      "step": 15413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7091991901397705,
+      "learning_rate": 4.592809939201692e-05,
+      "loss": 1.5751,
+      "step": 15414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9189236164093018,
+      "learning_rate": 4.592783505154639e-05,
+      "loss": 1.3686,
+      "step": 15415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1925766468048096,
+      "learning_rate": 4.592757071107587e-05,
+      "loss": 1.3562,
+      "step": 15416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.866569995880127,
+      "learning_rate": 4.592730637060535e-05,
+      "loss": 1.4557,
+      "step": 15417
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.192704916000366,
+      "learning_rate": 4.5927042030134814e-05,
+      "loss": 1.549,
+      "step": 15418
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.632671594619751,
+      "learning_rate": 4.592677768966429e-05,
+      "loss": 1.3592,
+      "step": 15419
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8075231313705444,
+      "learning_rate": 4.592651334919376e-05,
+      "loss": 1.4858,
+      "step": 15420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.46600079536438,
+      "learning_rate": 4.5926249008723235e-05,
+      "loss": 1.4294,
+      "step": 15421
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.072990894317627,
+      "learning_rate": 4.5925984668252715e-05,
+      "loss": 1.4551,
+      "step": 15422
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0004913806915283,
+      "learning_rate": 4.592572032778218e-05,
+      "loss": 1.4254,
+      "step": 15423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.769935965538025,
+      "learning_rate": 4.592545598731166e-05,
+      "loss": 1.3761,
+      "step": 15424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6234759092330933,
+      "learning_rate": 4.592519164684113e-05,
+      "loss": 1.3967,
+      "step": 15425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0066704750061035,
+      "learning_rate": 4.592492730637061e-05,
+      "loss": 1.4487,
+      "step": 15426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6365886926651,
+      "learning_rate": 4.592466296590008e-05,
+      "loss": 1.3365,
+      "step": 15427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.031498432159424,
+      "learning_rate": 4.5924398625429557e-05,
+      "loss": 1.5406,
+      "step": 15428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.120744228363037,
+      "learning_rate": 4.592413428495903e-05,
+      "loss": 1.4295,
+      "step": 15429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.816269874572754,
+      "learning_rate": 4.5923869944488504e-05,
+      "loss": 1.3465,
+      "step": 15430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.272575616836548,
+      "learning_rate": 4.592360560401798e-05,
+      "loss": 1.4832,
+      "step": 15431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7534900903701782,
+      "learning_rate": 4.592334126354745e-05,
+      "loss": 1.3435,
+      "step": 15432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8052666187286377,
+      "learning_rate": 4.5923076923076924e-05,
+      "loss": 1.3476,
+      "step": 15433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7541251182556152,
+      "learning_rate": 4.59228125826064e-05,
+      "loss": 1.5378,
+      "step": 15434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9571908712387085,
+      "learning_rate": 4.592254824213587e-05,
+      "loss": 1.3223,
+      "step": 15435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1510889530181885,
+      "learning_rate": 4.5922283901665345e-05,
+      "loss": 1.5002,
+      "step": 15436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2862536907196045,
+      "learning_rate": 4.5922019561194825e-05,
+      "loss": 1.4895,
+      "step": 15437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4363964796066284,
+      "learning_rate": 4.592175522072429e-05,
+      "loss": 1.3994,
+      "step": 15438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7270113229751587,
+      "learning_rate": 4.592149088025377e-05,
+      "loss": 1.1458,
+      "step": 15439
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5857621431350708,
+      "learning_rate": 4.592122653978324e-05,
+      "loss": 1.3796,
+      "step": 15440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7902450561523438,
+      "learning_rate": 4.592096219931272e-05,
+      "loss": 1.4322,
+      "step": 15441
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7086341381072998,
+      "learning_rate": 4.592069785884219e-05,
+      "loss": 1.3091,
+      "step": 15442
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.094780445098877,
+      "learning_rate": 4.592043351837167e-05,
+      "loss": 1.37,
+      "step": 15443
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.773547649383545,
+      "learning_rate": 4.592016917790114e-05,
+      "loss": 1.3044,
+      "step": 15444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6458345651626587,
+      "learning_rate": 4.5919904837430614e-05,
+      "loss": 1.2445,
+      "step": 15445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.254798173904419,
+      "learning_rate": 4.591964049696009e-05,
+      "loss": 1.4058,
+      "step": 15446
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.223245859146118,
+      "learning_rate": 4.591937615648956e-05,
+      "loss": 1.3653,
+      "step": 15447
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6748837232589722,
+      "learning_rate": 4.5919111816019035e-05,
+      "loss": 1.2555,
+      "step": 15448
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.06707763671875,
+      "learning_rate": 4.591884747554851e-05,
+      "loss": 1.5635,
+      "step": 15449
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3069608211517334,
+      "learning_rate": 4.591858313507798e-05,
+      "loss": 1.5394,
+      "step": 15450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4262547492980957,
+      "learning_rate": 4.5918318794607455e-05,
+      "loss": 1.4397,
+      "step": 15451
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8647074699401855,
+      "learning_rate": 4.5918054454136936e-05,
+      "loss": 1.4789,
+      "step": 15452
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.808510184288025,
+      "learning_rate": 4.59177901136664e-05,
+      "loss": 1.5832,
+      "step": 15453
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5168962478637695,
+      "learning_rate": 4.591752577319588e-05,
+      "loss": 1.2559,
+      "step": 15454
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0159590244293213,
+      "learning_rate": 4.591726143272535e-05,
+      "loss": 1.5726,
+      "step": 15455
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0712122917175293,
+      "learning_rate": 4.591699709225482e-05,
+      "loss": 1.4797,
+      "step": 15456
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7746962308883667,
+      "learning_rate": 4.5916732751784304e-05,
+      "loss": 1.7604,
+      "step": 15457
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4153517484664917,
+      "learning_rate": 4.591646841131377e-05,
+      "loss": 1.305,
+      "step": 15458
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.45265531539917,
+      "learning_rate": 4.591620407084325e-05,
+      "loss": 1.186,
+      "step": 15459
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.278688669204712,
+      "learning_rate": 4.591593973037272e-05,
+      "loss": 1.3401,
+      "step": 15460
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5833406448364258,
+      "learning_rate": 4.59156753899022e-05,
+      "loss": 1.2435,
+      "step": 15461
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.951055884361267,
+      "learning_rate": 4.591541104943167e-05,
+      "loss": 1.376,
+      "step": 15462
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.132085084915161,
+      "learning_rate": 4.5915146708961145e-05,
+      "loss": 1.6295,
+      "step": 15463
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9074037075042725,
+      "learning_rate": 4.591488236849062e-05,
+      "loss": 1.3659,
+      "step": 15464
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.956181526184082,
+      "learning_rate": 4.591461802802009e-05,
+      "loss": 1.3168,
+      "step": 15465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9752793312072754,
+      "learning_rate": 4.5914353687549566e-05,
+      "loss": 1.5546,
+      "step": 15466
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7550643682479858,
+      "learning_rate": 4.591408934707904e-05,
+      "loss": 1.483,
+      "step": 15467
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7364144325256348,
+      "learning_rate": 4.591382500660851e-05,
+      "loss": 1.3057,
+      "step": 15468
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.550176739692688,
+      "learning_rate": 4.5913560666137986e-05,
+      "loss": 1.3326,
+      "step": 15469
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.744594693183899,
+      "learning_rate": 4.591329632566746e-05,
+      "loss": 1.4687,
+      "step": 15470
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8896414041519165,
+      "learning_rate": 4.5913031985196934e-05,
+      "loss": 1.3909,
+      "step": 15471
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7694207429885864,
+      "learning_rate": 4.5912767644726414e-05,
+      "loss": 1.3187,
+      "step": 15472
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5357496738433838,
+      "learning_rate": 4.591250330425588e-05,
+      "loss": 2.07,
+      "step": 15473
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.925781011581421,
+      "learning_rate": 4.591223896378536e-05,
+      "loss": 1.2739,
+      "step": 15474
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.93535315990448,
+      "learning_rate": 4.591197462331483e-05,
+      "loss": 1.3432,
+      "step": 15475
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.50325345993042,
+      "learning_rate": 4.591171028284431e-05,
+      "loss": 1.5553,
+      "step": 15476
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8656823635101318,
+      "learning_rate": 4.591144594237378e-05,
+      "loss": 1.3663,
+      "step": 15477
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8569501638412476,
+      "learning_rate": 4.5911181601903255e-05,
+      "loss": 1.3201,
+      "step": 15478
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.654567837715149,
+      "learning_rate": 4.591091726143273e-05,
+      "loss": 1.4386,
+      "step": 15479
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.715020775794983,
+      "learning_rate": 4.59106529209622e-05,
+      "loss": 1.4812,
+      "step": 15480
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8994135856628418,
+      "learning_rate": 4.5910388580491676e-05,
+      "loss": 1.3931,
+      "step": 15481
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7680431604385376,
+      "learning_rate": 4.591012424002115e-05,
+      "loss": 1.5226,
+      "step": 15482
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.387861490249634,
+      "learning_rate": 4.590985989955062e-05,
+      "loss": 1.5525,
+      "step": 15483
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.06020450592041,
+      "learning_rate": 4.59095955590801e-05,
+      "loss": 1.5267,
+      "step": 15484
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.055332899093628,
+      "learning_rate": 4.590933121860957e-05,
+      "loss": 1.5842,
+      "step": 15485
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.140367269515991,
+      "learning_rate": 4.5909066878139044e-05,
+      "loss": 1.6793,
+      "step": 15486
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.808328628540039,
+      "learning_rate": 4.5908802537668524e-05,
+      "loss": 1.5758,
+      "step": 15487
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9822582006454468,
+      "learning_rate": 4.590853819719799e-05,
+      "loss": 1.4272,
+      "step": 15488
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.945492148399353,
+      "learning_rate": 4.590827385672747e-05,
+      "loss": 1.3832,
+      "step": 15489
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0713062286376953,
+      "learning_rate": 4.590800951625694e-05,
+      "loss": 1.424,
+      "step": 15490
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.022294044494629,
+      "learning_rate": 4.590774517578642e-05,
+      "loss": 1.5529,
+      "step": 15491
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.035714626312256,
+      "learning_rate": 4.590748083531589e-05,
+      "loss": 1.3119,
+      "step": 15492
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.197511672973633,
+      "learning_rate": 4.590721649484536e-05,
+      "loss": 1.3838,
+      "step": 15493
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8332629203796387,
+      "learning_rate": 4.590695215437484e-05,
+      "loss": 1.4811,
+      "step": 15494
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9753092527389526,
+      "learning_rate": 4.5906687813904306e-05,
+      "loss": 1.453,
+      "step": 15495
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7048392295837402,
+      "learning_rate": 4.5906423473433786e-05,
+      "loss": 1.2639,
+      "step": 15496
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.599040150642395,
+      "learning_rate": 4.590615913296326e-05,
+      "loss": 1.3578,
+      "step": 15497
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.783588171005249,
+      "learning_rate": 4.5905894792492734e-05,
+      "loss": 1.4227,
+      "step": 15498
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5514490604400635,
+      "learning_rate": 4.590563045202221e-05,
+      "loss": 1.414,
+      "step": 15499
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7166498899459839,
+      "learning_rate": 4.590536611155168e-05,
+      "loss": 1.4416,
+      "step": 15500
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5809051990509033,
+      "learning_rate": 4.5905101771081154e-05,
+      "loss": 1.3725,
+      "step": 15501
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8520565032958984,
+      "learning_rate": 4.590483743061063e-05,
+      "loss": 1.3876,
+      "step": 15502
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9016841650009155,
+      "learning_rate": 4.59045730901401e-05,
+      "loss": 1.4497,
+      "step": 15503
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8721963167190552,
+      "learning_rate": 4.5904308749669575e-05,
+      "loss": 1.3769,
+      "step": 15504
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5355366468429565,
+      "learning_rate": 4.590404440919905e-05,
+      "loss": 1.3045,
+      "step": 15505
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4455649852752686,
+      "learning_rate": 4.590378006872852e-05,
+      "loss": 1.2881,
+      "step": 15506
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4784106016159058,
+      "learning_rate": 4.5903515728258e-05,
+      "loss": 2.0918,
+      "step": 15507
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7043696641921997,
+      "learning_rate": 4.590325138778747e-05,
+      "loss": 1.1998,
+      "step": 15508
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0200135707855225,
+      "learning_rate": 4.590298704731695e-05,
+      "loss": 1.5502,
+      "step": 15509
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4166102409362793,
+      "learning_rate": 4.5902722706846416e-05,
+      "loss": 1.1913,
+      "step": 15510
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9908021688461304,
+      "learning_rate": 4.59024583663759e-05,
+      "loss": 2.1174,
+      "step": 15511
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8424533605575562,
+      "learning_rate": 4.590219402590537e-05,
+      "loss": 1.4085,
+      "step": 15512
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1332764625549316,
+      "learning_rate": 4.5901929685434844e-05,
+      "loss": 1.216,
+      "step": 15513
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4763294458389282,
+      "learning_rate": 4.590166534496432e-05,
+      "loss": 1.3132,
+      "step": 15514
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.083467960357666,
+      "learning_rate": 4.590140100449379e-05,
+      "loss": 1.3572,
+      "step": 15515
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.690003514289856,
+      "learning_rate": 4.5901136664023265e-05,
+      "loss": 1.4479,
+      "step": 15516
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1150128841400146,
+      "learning_rate": 4.590087232355274e-05,
+      "loss": 1.4726,
+      "step": 15517
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7155910730361938,
+      "learning_rate": 4.590060798308221e-05,
+      "loss": 1.4296,
+      "step": 15518
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6188085079193115,
+      "learning_rate": 4.5900343642611685e-05,
+      "loss": 1.1753,
+      "step": 15519
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6090940237045288,
+      "learning_rate": 4.590007930214116e-05,
+      "loss": 1.369,
+      "step": 15520
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1733460426330566,
+      "learning_rate": 4.589981496167063e-05,
+      "loss": 1.555,
+      "step": 15521
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6416155099868774,
+      "learning_rate": 4.589955062120011e-05,
+      "loss": 1.2866,
+      "step": 15522
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5979220867156982,
+      "learning_rate": 4.589928628072958e-05,
+      "loss": 1.3206,
+      "step": 15523
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.590039849281311,
+      "learning_rate": 4.589902194025906e-05,
+      "loss": 1.1923,
+      "step": 15524
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6738200187683105,
+      "learning_rate": 4.589875759978853e-05,
+      "loss": 1.3194,
+      "step": 15525
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.594711184501648,
+      "learning_rate": 4.589849325931801e-05,
+      "loss": 1.304,
+      "step": 15526
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7220615148544312,
+      "learning_rate": 4.589822891884748e-05,
+      "loss": 1.4477,
+      "step": 15527
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7903416156768799,
+      "learning_rate": 4.589796457837695e-05,
+      "loss": 1.3437,
+      "step": 15528
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.65766441822052,
+      "learning_rate": 4.589770023790643e-05,
+      "loss": 1.6649,
+      "step": 15529
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.042386770248413,
+      "learning_rate": 4.5897435897435895e-05,
+      "loss": 1.5,
+      "step": 15530
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.16827654838562,
+      "learning_rate": 4.5897171556965375e-05,
+      "loss": 1.594,
+      "step": 15531
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9909785985946655,
+      "learning_rate": 4.589690721649485e-05,
+      "loss": 1.3983,
+      "step": 15532
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9946502447128296,
+      "learning_rate": 4.589664287602432e-05,
+      "loss": 1.4743,
+      "step": 15533
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4578790664672852,
+      "learning_rate": 4.5896378535553796e-05,
+      "loss": 1.3192,
+      "step": 15534
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0969154834747314,
+      "learning_rate": 4.589611419508327e-05,
+      "loss": 1.4205,
+      "step": 15535
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9545691013336182,
+      "learning_rate": 4.589584985461274e-05,
+      "loss": 1.4988,
+      "step": 15536
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9093247652053833,
+      "learning_rate": 4.5895585514142216e-05,
+      "loss": 1.3233,
+      "step": 15537
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.333265781402588,
+      "learning_rate": 4.589532117367169e-05,
+      "loss": 1.3519,
+      "step": 15538
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.188565731048584,
+      "learning_rate": 4.5895056833201164e-05,
+      "loss": 1.5316,
+      "step": 15539
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.694053053855896,
+      "learning_rate": 4.589479249273064e-05,
+      "loss": 1.4206,
+      "step": 15540
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4358766078948975,
+      "learning_rate": 4.589452815226011e-05,
+      "loss": 1.3257,
+      "step": 15541
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.183907985687256,
+      "learning_rate": 4.589426381178959e-05,
+      "loss": 1.4232,
+      "step": 15542
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8015265464782715,
+      "learning_rate": 4.589399947131906e-05,
+      "loss": 1.2739,
+      "step": 15543
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4294885396957397,
+      "learning_rate": 4.589373513084854e-05,
+      "loss": 1.2666,
+      "step": 15544
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.469130277633667,
+      "learning_rate": 4.5893470790378005e-05,
+      "loss": 1.2087,
+      "step": 15545
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.249488592147827,
+      "learning_rate": 4.5893206449907485e-05,
+      "loss": 1.3415,
+      "step": 15546
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8538029193878174,
+      "learning_rate": 4.589294210943696e-05,
+      "loss": 1.4958,
+      "step": 15547
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6211776733398438,
+      "learning_rate": 4.589267776896643e-05,
+      "loss": 1.2626,
+      "step": 15548
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.470531702041626,
+      "learning_rate": 4.5892413428495906e-05,
+      "loss": 1.4044,
+      "step": 15549
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0361976623535156,
+      "learning_rate": 4.589214908802538e-05,
+      "loss": 1.4114,
+      "step": 15550
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8603851795196533,
+      "learning_rate": 4.589188474755485e-05,
+      "loss": 2.3714,
+      "step": 15551
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0789148807525635,
+      "learning_rate": 4.589162040708433e-05,
+      "loss": 1.47,
+      "step": 15552
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9890975952148438,
+      "learning_rate": 4.58913560666138e-05,
+      "loss": 1.4036,
+      "step": 15553
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.59356427192688,
+      "learning_rate": 4.5891091726143274e-05,
+      "loss": 1.2646,
+      "step": 15554
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9884995222091675,
+      "learning_rate": 4.589082738567275e-05,
+      "loss": 1.5199,
+      "step": 15555
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8889662027359009,
+      "learning_rate": 4.589056304520222e-05,
+      "loss": 1.4296,
+      "step": 15556
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2171523571014404,
+      "learning_rate": 4.58902987047317e-05,
+      "loss": 1.3986,
+      "step": 15557
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5339467525482178,
+      "learning_rate": 4.589003436426117e-05,
+      "loss": 1.2315,
+      "step": 15558
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.219304084777832,
+      "learning_rate": 4.588977002379065e-05,
+      "loss": 1.5667,
+      "step": 15559
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7865676879882812,
+      "learning_rate": 4.5889505683320115e-05,
+      "loss": 1.4444,
+      "step": 15560
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2155425548553467,
+      "learning_rate": 4.5889241342849596e-05,
+      "loss": 1.7725,
+      "step": 15561
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8665889501571655,
+      "learning_rate": 4.588897700237907e-05,
+      "loss": 1.3844,
+      "step": 15562
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.682104468345642,
+      "learning_rate": 4.588871266190854e-05,
+      "loss": 1.3282,
+      "step": 15563
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1653192043304443,
+      "learning_rate": 4.5888448321438016e-05,
+      "loss": 1.5753,
+      "step": 15564
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6159203052520752,
+      "learning_rate": 4.588818398096748e-05,
+      "loss": 1.3804,
+      "step": 15565
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.906407117843628,
+      "learning_rate": 4.5887919640496964e-05,
+      "loss": 1.4726,
+      "step": 15566
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6965595483779907,
+      "learning_rate": 4.588765530002644e-05,
+      "loss": 1.3227,
+      "step": 15567
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8136252164840698,
+      "learning_rate": 4.588739095955591e-05,
+      "loss": 1.3353,
+      "step": 15568
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.796600103378296,
+      "learning_rate": 4.5887126619085384e-05,
+      "loss": 1.3017,
+      "step": 15569
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6639931201934814,
+      "learning_rate": 4.588686227861486e-05,
+      "loss": 1.357,
+      "step": 15570
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8493719100952148,
+      "learning_rate": 4.588659793814433e-05,
+      "loss": 1.4582,
+      "step": 15571
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5877848863601685,
+      "learning_rate": 4.588633359767381e-05,
+      "loss": 1.4037,
+      "step": 15572
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5903652906417847,
+      "learning_rate": 4.588606925720328e-05,
+      "loss": 1.3349,
+      "step": 15573
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5661554336547852,
+      "learning_rate": 4.588580491673275e-05,
+      "loss": 1.3469,
+      "step": 15574
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9705750942230225,
+      "learning_rate": 4.5885540576262226e-05,
+      "loss": 1.5073,
+      "step": 15575
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.61650550365448,
+      "learning_rate": 4.58852762357917e-05,
+      "loss": 1.3847,
+      "step": 15576
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0114352703094482,
+      "learning_rate": 4.588501189532118e-05,
+      "loss": 1.3701,
+      "step": 15577
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.272656202316284,
+      "learning_rate": 4.5884747554850646e-05,
+      "loss": 1.3653,
+      "step": 15578
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6652432680130005,
+      "learning_rate": 4.588448321438013e-05,
+      "loss": 1.3697,
+      "step": 15579
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.906448483467102,
+      "learning_rate": 4.5884218873909594e-05,
+      "loss": 1.2751,
+      "step": 15580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9837161302566528,
+      "learning_rate": 4.5883954533439074e-05,
+      "loss": 1.304,
+      "step": 15581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8166996240615845,
+      "learning_rate": 4.588369019296855e-05,
+      "loss": 1.3291,
+      "step": 15582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5194400548934937,
+      "learning_rate": 4.588342585249802e-05,
+      "loss": 1.4441,
+      "step": 15583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7238898277282715,
+      "learning_rate": 4.5883161512027495e-05,
+      "loss": 1.4195,
+      "step": 15584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7078113555908203,
+      "learning_rate": 4.588289717155697e-05,
+      "loss": 1.5089,
+      "step": 15585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.91016685962677,
+      "learning_rate": 4.588263283108644e-05,
+      "loss": 1.3862,
+      "step": 15586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.124560832977295,
+      "learning_rate": 4.5882368490615915e-05,
+      "loss": 1.4519,
+      "step": 15587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.509495973587036,
+      "learning_rate": 4.588210415014539e-05,
+      "loss": 1.7761,
+      "step": 15588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7198266983032227,
+      "learning_rate": 4.588183980967486e-05,
+      "loss": 1.3546,
+      "step": 15589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9860658645629883,
+      "learning_rate": 4.5881575469204336e-05,
+      "loss": 1.3178,
+      "step": 15590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.062567949295044,
+      "learning_rate": 4.588131112873381e-05,
+      "loss": 1.3604,
+      "step": 15591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8759585618972778,
+      "learning_rate": 4.588104678826329e-05,
+      "loss": 1.4293,
+      "step": 15592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7166049480438232,
+      "learning_rate": 4.588078244779276e-05,
+      "loss": 1.4654,
+      "step": 15593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.506837248802185,
+      "learning_rate": 4.588051810732224e-05,
+      "loss": 1.3827,
+      "step": 15594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9128894805908203,
+      "learning_rate": 4.5880253766851704e-05,
+      "loss": 1.6834,
+      "step": 15595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8565006256103516,
+      "learning_rate": 4.5879989426381184e-05,
+      "loss": 1.425,
+      "step": 15596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5112971067428589,
+      "learning_rate": 4.587972508591066e-05,
+      "loss": 1.3224,
+      "step": 15597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.007695198059082,
+      "learning_rate": 4.587946074544013e-05,
+      "loss": 1.4189,
+      "step": 15598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7559611797332764,
+      "learning_rate": 4.5879196404969605e-05,
+      "loss": 1.3484,
+      "step": 15599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.684478282928467,
+      "learning_rate": 4.587893206449907e-05,
+      "loss": 1.5199,
+      "step": 15600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0763704776763916,
+      "learning_rate": 4.587866772402855e-05,
+      "loss": 1.5449,
+      "step": 15601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6974660158157349,
+      "learning_rate": 4.5878403383558026e-05,
+      "loss": 1.4947,
+      "step": 15602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1765754222869873,
+      "learning_rate": 4.58781390430875e-05,
+      "loss": 1.2475,
+      "step": 15603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4219985008239746,
+      "learning_rate": 4.587787470261697e-05,
+      "loss": 1.4524,
+      "step": 15604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5826034545898438,
+      "learning_rate": 4.5877610362146446e-05,
+      "loss": 1.2688,
+      "step": 15605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3088085651397705,
+      "learning_rate": 4.587734602167592e-05,
+      "loss": 1.2517,
+      "step": 15606
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2628064155578613,
+      "learning_rate": 4.58770816812054e-05,
+      "loss": 1.4081,
+      "step": 15607
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5674782991409302,
+      "learning_rate": 4.587681734073487e-05,
+      "loss": 1.3255,
+      "step": 15608
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.718025803565979,
+      "learning_rate": 4.587655300026434e-05,
+      "loss": 1.4426,
+      "step": 15609
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7912094593048096,
+      "learning_rate": 4.5876288659793814e-05,
+      "loss": 1.4098,
+      "step": 15610
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8013778924942017,
+      "learning_rate": 4.587602431932329e-05,
+      "loss": 1.532,
+      "step": 15611
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8949490785598755,
+      "learning_rate": 4.587575997885277e-05,
+      "loss": 1.3165,
+      "step": 15612
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4943403005599976,
+      "learning_rate": 4.5875495638382235e-05,
+      "loss": 1.1707,
+      "step": 15613
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.332530975341797,
+      "learning_rate": 4.5875231297911715e-05,
+      "loss": 1.4331,
+      "step": 15614
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8609980344772339,
+      "learning_rate": 4.587496695744118e-05,
+      "loss": 1.5657,
+      "step": 15615
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7951055765151978,
+      "learning_rate": 4.587470261697066e-05,
+      "loss": 1.3344,
+      "step": 15616
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7055132389068604,
+      "learning_rate": 4.5874438276500136e-05,
+      "loss": 1.2236,
+      "step": 15617
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.742438554763794,
+      "learning_rate": 4.587417393602961e-05,
+      "loss": 1.3052,
+      "step": 15618
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2963736057281494,
+      "learning_rate": 4.587390959555908e-05,
+      "loss": 2.3603,
+      "step": 15619
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5644046068191528,
+      "learning_rate": 4.587364525508856e-05,
+      "loss": 1.2716,
+      "step": 15620
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8748284578323364,
+      "learning_rate": 4.587338091461803e-05,
+      "loss": 1.5013,
+      "step": 15621
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7179957628250122,
+      "learning_rate": 4.5873116574147504e-05,
+      "loss": 1.4487,
+      "step": 15622
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7467948198318481,
+      "learning_rate": 4.587285223367698e-05,
+      "loss": 1.2008,
+      "step": 15623
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3474444150924683,
+      "learning_rate": 4.587258789320645e-05,
+      "loss": 1.3724,
+      "step": 15624
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.377238154411316,
+      "learning_rate": 4.5872323552735925e-05,
+      "loss": 2.3394,
+      "step": 15625
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9904141426086426,
+      "learning_rate": 4.58720592122654e-05,
+      "loss": 1.4069,
+      "step": 15626
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6135543584823608,
+      "learning_rate": 4.587179487179488e-05,
+      "loss": 1.3575,
+      "step": 15627
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8651785850524902,
+      "learning_rate": 4.5871530531324345e-05,
+      "loss": 1.3739,
+      "step": 15628
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8703855276107788,
+      "learning_rate": 4.5871266190853826e-05,
+      "loss": 1.4109,
+      "step": 15629
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.576311469078064,
+      "learning_rate": 4.587100185038329e-05,
+      "loss": 1.3615,
+      "step": 15630
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5498528480529785,
+      "learning_rate": 4.587073750991277e-05,
+      "loss": 1.4691,
+      "step": 15631
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.823604106903076,
+      "learning_rate": 4.5870473169442246e-05,
+      "loss": 1.4012,
+      "step": 15632
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.656758189201355,
+      "learning_rate": 4.587020882897172e-05,
+      "loss": 1.3413,
+      "step": 15633
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6859122514724731,
+      "learning_rate": 4.5869944488501193e-05,
+      "loss": 1.4248,
+      "step": 15634
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0594704151153564,
+      "learning_rate": 4.586968014803067e-05,
+      "loss": 1.3792,
+      "step": 15635
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4927055835723877,
+      "learning_rate": 4.586941580756014e-05,
+      "loss": 1.2938,
+      "step": 15636
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6963316202163696,
+      "learning_rate": 4.5869151467089614e-05,
+      "loss": 1.5075,
+      "step": 15637
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7809802293777466,
+      "learning_rate": 4.586888712661909e-05,
+      "loss": 1.3564,
+      "step": 15638
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.817790150642395,
+      "learning_rate": 4.586862278614856e-05,
+      "loss": 1.3401,
+      "step": 15639
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.819351673126221,
+      "learning_rate": 4.5868358445678035e-05,
+      "loss": 1.5335,
+      "step": 15640
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8969401121139526,
+      "learning_rate": 4.586809410520751e-05,
+      "loss": 1.3415,
+      "step": 15641
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5632147789001465,
+      "learning_rate": 4.586782976473699e-05,
+      "loss": 1.372,
+      "step": 15642
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5635449886322021,
+      "learning_rate": 4.5867565424266456e-05,
+      "loss": 1.3328,
+      "step": 15643
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8179619312286377,
+      "learning_rate": 4.5867301083795936e-05,
+      "loss": 1.6341,
+      "step": 15644
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8248804807662964,
+      "learning_rate": 4.58670367433254e-05,
+      "loss": 1.5303,
+      "step": 15645
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.783109188079834,
+      "learning_rate": 4.5866772402854876e-05,
+      "loss": 1.3572,
+      "step": 15646
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2067368030548096,
+      "learning_rate": 4.586650806238436e-05,
+      "loss": 1.4724,
+      "step": 15647
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.121504783630371,
+      "learning_rate": 4.5866243721913823e-05,
+      "loss": 1.6368,
+      "step": 15648
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.958003044128418,
+      "learning_rate": 4.5865979381443304e-05,
+      "loss": 2.1009,
+      "step": 15649
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5338010787963867,
+      "learning_rate": 4.586571504097277e-05,
+      "loss": 1.3202,
+      "step": 15650
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9349201917648315,
+      "learning_rate": 4.586545070050225e-05,
+      "loss": 1.4617,
+      "step": 15651
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.719209909439087,
+      "learning_rate": 4.5865186360031724e-05,
+      "loss": 1.4432,
+      "step": 15652
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7430810928344727,
+      "learning_rate": 4.58649220195612e-05,
+      "loss": 1.478,
+      "step": 15653
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6645681858062744,
+      "learning_rate": 4.586465767909067e-05,
+      "loss": 1.1532,
+      "step": 15654
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9149905443191528,
+      "learning_rate": 4.5864393338620145e-05,
+      "loss": 1.5434,
+      "step": 15655
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5073422193527222,
+      "learning_rate": 4.586412899814962e-05,
+      "loss": 1.3764,
+      "step": 15656
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1030499935150146,
+      "learning_rate": 4.586386465767909e-05,
+      "loss": 1.3245,
+      "step": 15657
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5995893478393555,
+      "learning_rate": 4.5863600317208566e-05,
+      "loss": 1.4532,
+      "step": 15658
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5884696245193481,
+      "learning_rate": 4.586333597673804e-05,
+      "loss": 1.2365,
+      "step": 15659
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0308351516723633,
+      "learning_rate": 4.586307163626751e-05,
+      "loss": 1.4201,
+      "step": 15660
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2027297019958496,
+      "learning_rate": 4.586280729579699e-05,
+      "loss": 1.4017,
+      "step": 15661
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4285281896591187,
+      "learning_rate": 4.586254295532647e-05,
+      "loss": 1.731,
+      "step": 15662
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8877846002578735,
+      "learning_rate": 4.5862278614855934e-05,
+      "loss": 1.4714,
+      "step": 15663
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.132652521133423,
+      "learning_rate": 4.5862014274385414e-05,
+      "loss": 1.3047,
+      "step": 15664
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4519679546356201,
+      "learning_rate": 4.586174993391488e-05,
+      "loss": 1.3448,
+      "step": 15665
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.598839282989502,
+      "learning_rate": 4.586148559344436e-05,
+      "loss": 1.1322,
+      "step": 15666
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0549540519714355,
+      "learning_rate": 4.5861221252973835e-05,
+      "loss": 1.4924,
+      "step": 15667
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5437211990356445,
+      "learning_rate": 4.586095691250331e-05,
+      "loss": 1.2892,
+      "step": 15668
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.019787073135376,
+      "learning_rate": 4.586069257203278e-05,
+      "loss": 1.509,
+      "step": 15669
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.740479588508606,
+      "learning_rate": 4.5860428231562256e-05,
+      "loss": 1.3689,
+      "step": 15670
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2331316471099854,
+      "learning_rate": 4.586016389109173e-05,
+      "loss": 1.4856,
+      "step": 15671
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7364834547042847,
+      "learning_rate": 4.58598995506212e-05,
+      "loss": 1.4097,
+      "step": 15672
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6124625205993652,
+      "learning_rate": 4.5859635210150676e-05,
+      "loss": 2.3023,
+      "step": 15673
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7587859630584717,
+      "learning_rate": 4.585937086968015e-05,
+      "loss": 1.4588,
+      "step": 15674
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9853267669677734,
+      "learning_rate": 4.5859106529209623e-05,
+      "loss": 1.3949,
+      "step": 15675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9474903345108032,
+      "learning_rate": 4.58588421887391e-05,
+      "loss": 1.3439,
+      "step": 15676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5807487964630127,
+      "learning_rate": 4.585857784826858e-05,
+      "loss": 1.3942,
+      "step": 15677
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7166279554367065,
+      "learning_rate": 4.5858313507798044e-05,
+      "loss": 1.5038,
+      "step": 15678
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5849404335021973,
+      "learning_rate": 4.5858049167327524e-05,
+      "loss": 1.3077,
+      "step": 15679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3966853618621826,
+      "learning_rate": 4.585778482685699e-05,
+      "loss": 1.3229,
+      "step": 15680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5299173593521118,
+      "learning_rate": 4.585752048638647e-05,
+      "loss": 1.4128,
+      "step": 15681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8116689920425415,
+      "learning_rate": 4.5857256145915945e-05,
+      "loss": 1.3738,
+      "step": 15682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9727163314819336,
+      "learning_rate": 4.585699180544541e-05,
+      "loss": 1.4346,
+      "step": 15683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7317293882369995,
+      "learning_rate": 4.585672746497489e-05,
+      "loss": 1.4138,
+      "step": 15684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6622610092163086,
+      "learning_rate": 4.585646312450436e-05,
+      "loss": 1.3519,
+      "step": 15685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7540310621261597,
+      "learning_rate": 4.585619878403384e-05,
+      "loss": 1.3408,
+      "step": 15686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5233943462371826,
+      "learning_rate": 4.585593444356331e-05,
+      "loss": 1.4559,
+      "step": 15687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5191432237625122,
+      "learning_rate": 4.5855670103092787e-05,
+      "loss": 1.3384,
+      "step": 15688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0910744667053223,
+      "learning_rate": 4.585540576262226e-05,
+      "loss": 1.5085,
+      "step": 15689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8030223846435547,
+      "learning_rate": 4.5855141422151734e-05,
+      "loss": 1.3098,
+      "step": 15690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9232947826385498,
+      "learning_rate": 4.585487708168121e-05,
+      "loss": 1.4624,
+      "step": 15691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4931451082229614,
+      "learning_rate": 4.585461274121068e-05,
+      "loss": 1.2818,
+      "step": 15692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.407517194747925,
+      "learning_rate": 4.5854348400740154e-05,
+      "loss": 1.2961,
+      "step": 15693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0204501152038574,
+      "learning_rate": 4.585408406026963e-05,
+      "loss": 1.4877,
+      "step": 15694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7861677408218384,
+      "learning_rate": 4.58538197197991e-05,
+      "loss": 1.3364,
+      "step": 15695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.960034966468811,
+      "learning_rate": 4.5853555379328575e-05,
+      "loss": 1.5193,
+      "step": 15696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6028540134429932,
+      "learning_rate": 4.5853291038858056e-05,
+      "loss": 1.4634,
+      "step": 15697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7990425825119019,
+      "learning_rate": 4.585302669838752e-05,
+      "loss": 1.4836,
+      "step": 15698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6169235706329346,
+      "learning_rate": 4.5852762357917e-05,
+      "loss": 1.2797,
+      "step": 15699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.55448579788208,
+      "learning_rate": 4.585249801744647e-05,
+      "loss": 1.3192,
+      "step": 15700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6679900884628296,
+      "learning_rate": 4.585223367697595e-05,
+      "loss": 1.4065,
+      "step": 15701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.836082935333252,
+      "learning_rate": 4.585196933650542e-05,
+      "loss": 1.5197,
+      "step": 15702
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.058134078979492,
+      "learning_rate": 4.58517049960349e-05,
+      "loss": 1.2652,
+      "step": 15703
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8115369081497192,
+      "learning_rate": 4.585144065556437e-05,
+      "loss": 1.4396,
+      "step": 15704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6297844648361206,
+      "learning_rate": 4.5851176315093844e-05,
+      "loss": 1.4662,
+      "step": 15705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.257067918777466,
+      "learning_rate": 4.585091197462332e-05,
+      "loss": 1.3721,
+      "step": 15706
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9725980758666992,
+      "learning_rate": 4.585064763415279e-05,
+      "loss": 1.495,
+      "step": 15707
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6620432138442993,
+      "learning_rate": 4.5850383293682265e-05,
+      "loss": 1.274,
+      "step": 15708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5732455253601074,
+      "learning_rate": 4.585011895321174e-05,
+      "loss": 1.391,
+      "step": 15709
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0805578231811523,
+      "learning_rate": 4.584985461274121e-05,
+      "loss": 1.479,
+      "step": 15710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8527352809906006,
+      "learning_rate": 4.5849590272270686e-05,
+      "loss": 1.4026,
+      "step": 15711
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.787176251411438,
+      "learning_rate": 4.5849325931800166e-05,
+      "loss": 1.5066,
+      "step": 15712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5639424324035645,
+      "learning_rate": 4.584906159132963e-05,
+      "loss": 1.4647,
+      "step": 15713
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.315746307373047,
+      "learning_rate": 4.584879725085911e-05,
+      "loss": 1.4949,
+      "step": 15714
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5095463991165161,
+      "learning_rate": 4.584853291038858e-05,
+      "loss": 1.3333,
+      "step": 15715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.794292688369751,
+      "learning_rate": 4.584826856991806e-05,
+      "loss": 1.3846,
+      "step": 15716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1312756538391113,
+      "learning_rate": 4.5848004229447534e-05,
+      "loss": 1.4432,
+      "step": 15717
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5482196807861328,
+      "learning_rate": 4.5847739888977e-05,
+      "loss": 1.295,
+      "step": 15718
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6768527030944824,
+      "learning_rate": 4.584747554850648e-05,
+      "loss": 1.3629,
+      "step": 15719
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6118186712265015,
+      "learning_rate": 4.584721120803595e-05,
+      "loss": 1.449,
+      "step": 15720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8621231317520142,
+      "learning_rate": 4.584694686756543e-05,
+      "loss": 1.3012,
+      "step": 15721
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4481024742126465,
+      "learning_rate": 4.58466825270949e-05,
+      "loss": 1.492,
+      "step": 15722
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7792474031448364,
+      "learning_rate": 4.5846418186624375e-05,
+      "loss": 1.3484,
+      "step": 15723
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.062168598175049,
+      "learning_rate": 4.584615384615385e-05,
+      "loss": 1.5059,
+      "step": 15724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9332493543624878,
+      "learning_rate": 4.584588950568332e-05,
+      "loss": 1.4989,
+      "step": 15725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3521332740783691,
+      "learning_rate": 4.5845625165212796e-05,
+      "loss": 1.3569,
+      "step": 15726
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6590322256088257,
+      "learning_rate": 4.584536082474227e-05,
+      "loss": 1.3778,
+      "step": 15727
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9104329347610474,
+      "learning_rate": 4.584509648427174e-05,
+      "loss": 1.4435,
+      "step": 15728
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4712350368499756,
+      "learning_rate": 4.5844832143801217e-05,
+      "loss": 1.3365,
+      "step": 15729
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1093811988830566,
+      "learning_rate": 4.584456780333069e-05,
+      "loss": 1.3531,
+      "step": 15730
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8608149290084839,
+      "learning_rate": 4.5844303462860164e-05,
+      "loss": 1.4463,
+      "step": 15731
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7871568202972412,
+      "learning_rate": 4.5844039122389644e-05,
+      "loss": 1.3431,
+      "step": 15732
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6524630784988403,
+      "learning_rate": 4.584377478191911e-05,
+      "loss": 1.4556,
+      "step": 15733
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1573493480682373,
+      "learning_rate": 4.584351044144859e-05,
+      "loss": 2.3032,
+      "step": 15734
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9006130695343018,
+      "learning_rate": 4.584324610097806e-05,
+      "loss": 1.4443,
+      "step": 15735
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8333165645599365,
+      "learning_rate": 4.584298176050754e-05,
+      "loss": 1.414,
+      "step": 15736
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.02683687210083,
+      "learning_rate": 4.584271742003701e-05,
+      "loss": 1.5332,
+      "step": 15737
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.804791808128357,
+      "learning_rate": 4.5842453079566485e-05,
+      "loss": 1.3069,
+      "step": 15738
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5983855724334717,
+      "learning_rate": 4.584218873909596e-05,
+      "loss": 1.3711,
+      "step": 15739
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.200342893600464,
+      "learning_rate": 4.584192439862543e-05,
+      "loss": 1.4305,
+      "step": 15740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5507919788360596,
+      "learning_rate": 4.5841660058154906e-05,
+      "loss": 1.3031,
+      "step": 15741
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.42069935798645,
+      "learning_rate": 4.584139571768438e-05,
+      "loss": 1.5688,
+      "step": 15742
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8509833812713623,
+      "learning_rate": 4.584113137721385e-05,
+      "loss": 1.5151,
+      "step": 15743
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4973710775375366,
+      "learning_rate": 4.584086703674333e-05,
+      "loss": 1.2609,
+      "step": 15744
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.653113842010498,
+      "learning_rate": 4.58406026962728e-05,
+      "loss": 1.1294,
+      "step": 15745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5931440591812134,
+      "learning_rate": 4.5840338355802274e-05,
+      "loss": 1.3607,
+      "step": 15746
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.761478304862976,
+      "learning_rate": 4.5840074015331754e-05,
+      "loss": 1.4169,
+      "step": 15747
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6018049716949463,
+      "learning_rate": 4.583980967486122e-05,
+      "loss": 1.3979,
+      "step": 15748
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.772646188735962,
+      "learning_rate": 4.58395453343907e-05,
+      "loss": 1.2612,
+      "step": 15749
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2607454061508179,
+      "learning_rate": 4.583928099392017e-05,
+      "loss": 1.2542,
+      "step": 15750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5436526536941528,
+      "learning_rate": 4.583901665344965e-05,
+      "loss": 1.4386,
+      "step": 15751
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.240281581878662,
+      "learning_rate": 4.583875231297912e-05,
+      "loss": 1.6398,
+      "step": 15752
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7071378231048584,
+      "learning_rate": 4.5838487972508596e-05,
+      "loss": 1.4183,
+      "step": 15753
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.630528450012207,
+      "learning_rate": 4.583822363203807e-05,
+      "loss": 1.2986,
+      "step": 15754
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3328132629394531,
+      "learning_rate": 4.5837959291567536e-05,
+      "loss": 1.2314,
+      "step": 15755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.778775691986084,
+      "learning_rate": 4.5837694951097017e-05,
+      "loss": 1.4289,
+      "step": 15756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6612963676452637,
+      "learning_rate": 4.583743061062649e-05,
+      "loss": 1.4815,
+      "step": 15757
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3750760555267334,
+      "learning_rate": 4.5837166270155964e-05,
+      "loss": 1.7693,
+      "step": 15758
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.083576202392578,
+      "learning_rate": 4.583690192968544e-05,
+      "loss": 1.5979,
+      "step": 15759
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9692775011062622,
+      "learning_rate": 4.583663758921491e-05,
+      "loss": 1.6114,
+      "step": 15760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9313596487045288,
+      "learning_rate": 4.5836373248744384e-05,
+      "loss": 1.4884,
+      "step": 15761
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.773785948753357,
+      "learning_rate": 4.5836108908273865e-05,
+      "loss": 1.4198,
+      "step": 15762
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8039063215255737,
+      "learning_rate": 4.583584456780333e-05,
+      "loss": 1.7205,
+      "step": 15763
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6620533466339111,
+      "learning_rate": 4.5835580227332805e-05,
+      "loss": 1.3592,
+      "step": 15764
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.001307725906372,
+      "learning_rate": 4.583531588686228e-05,
+      "loss": 1.5886,
+      "step": 15765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5227551460266113,
+      "learning_rate": 4.583505154639175e-05,
+      "loss": 1.5023,
+      "step": 15766
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4513832330703735,
+      "learning_rate": 4.583478720592123e-05,
+      "loss": 1.2025,
+      "step": 15767
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1933822631835938,
+      "learning_rate": 4.58345228654507e-05,
+      "loss": 1.6252,
+      "step": 15768
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7377514839172363,
+      "learning_rate": 4.583425852498018e-05,
+      "loss": 1.4442,
+      "step": 15769
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0063939094543457,
+      "learning_rate": 4.5833994184509647e-05,
+      "loss": 1.4797,
+      "step": 15770
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7541446685791016,
+      "learning_rate": 4.583372984403913e-05,
+      "loss": 1.495,
+      "step": 15771
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5280793905258179,
+      "learning_rate": 4.58334655035686e-05,
+      "loss": 1.2475,
+      "step": 15772
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.682959794998169,
+      "learning_rate": 4.5833201163098074e-05,
+      "loss": 1.7189,
+      "step": 15773
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4142541885375977,
+      "learning_rate": 4.583293682262755e-05,
+      "loss": 1.5296,
+      "step": 15774
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.842722773551941,
+      "learning_rate": 4.583267248215702e-05,
+      "loss": 1.4167,
+      "step": 15775
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.407755374908447,
+      "learning_rate": 4.5832408141686495e-05,
+      "loss": 1.4771,
+      "step": 15776
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7653554677963257,
+      "learning_rate": 4.583214380121597e-05,
+      "loss": 1.3347,
+      "step": 15777
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5171797275543213,
+      "learning_rate": 4.583187946074544e-05,
+      "loss": 1.3873,
+      "step": 15778
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.41636323928833,
+      "learning_rate": 4.5831615120274915e-05,
+      "loss": 1.2743,
+      "step": 15779
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7430497407913208,
+      "learning_rate": 4.583135077980439e-05,
+      "loss": 2.1361,
+      "step": 15780
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0456714630126953,
+      "learning_rate": 4.583108643933386e-05,
+      "loss": 1.4433,
+      "step": 15781
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7517857551574707,
+      "learning_rate": 4.583082209886334e-05,
+      "loss": 1.4889,
+      "step": 15782
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7731066942214966,
+      "learning_rate": 4.583055775839281e-05,
+      "loss": 1.3639,
+      "step": 15783
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4555985927581787,
+      "learning_rate": 4.583029341792229e-05,
+      "loss": 1.3226,
+      "step": 15784
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7844058275222778,
+      "learning_rate": 4.583002907745176e-05,
+      "loss": 1.5641,
+      "step": 15785
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6895369291305542,
+      "learning_rate": 4.582976473698124e-05,
+      "loss": 1.3259,
+      "step": 15786
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5122687816619873,
+      "learning_rate": 4.582950039651071e-05,
+      "loss": 1.2739,
+      "step": 15787
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.567487359046936,
+      "learning_rate": 4.5829236056040184e-05,
+      "loss": 1.3737,
+      "step": 15788
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6873259544372559,
+      "learning_rate": 4.582897171556966e-05,
+      "loss": 1.3427,
+      "step": 15789
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5782418251037598,
+      "learning_rate": 4.5828707375099125e-05,
+      "loss": 1.3176,
+      "step": 15790
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8476042747497559,
+      "learning_rate": 4.5828443034628605e-05,
+      "loss": 1.4492,
+      "step": 15791
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9334408044815063,
+      "learning_rate": 4.582817869415808e-05,
+      "loss": 1.5642,
+      "step": 15792
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.700023889541626,
+      "learning_rate": 4.582791435368755e-05,
+      "loss": 1.3016,
+      "step": 15793
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4671622514724731,
+      "learning_rate": 4.5827650013217026e-05,
+      "loss": 1.4185,
+      "step": 15794
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9392693042755127,
+      "learning_rate": 4.58273856727465e-05,
+      "loss": 1.4699,
+      "step": 15795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.70673668384552,
+      "learning_rate": 4.582712133227597e-05,
+      "loss": 1.3024,
+      "step": 15796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5400716066360474,
+      "learning_rate": 4.582685699180545e-05,
+      "loss": 1.4369,
+      "step": 15797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0610742568969727,
+      "learning_rate": 4.582659265133492e-05,
+      "loss": 1.3243,
+      "step": 15798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1636064052581787,
+      "learning_rate": 4.58263283108644e-05,
+      "loss": 1.4578,
+      "step": 15799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4723783731460571,
+      "learning_rate": 4.582606397039387e-05,
+      "loss": 1.8589,
+      "step": 15800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7107049226760864,
+      "learning_rate": 4.582579962992334e-05,
+      "loss": 1.3659,
+      "step": 15801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4810519218444824,
+      "learning_rate": 4.582553528945282e-05,
+      "loss": 1.1774,
+      "step": 15802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7457778453826904,
+      "learning_rate": 4.582527094898229e-05,
+      "loss": 1.8942,
+      "step": 15803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7927929162979126,
+      "learning_rate": 4.582500660851177e-05,
+      "loss": 1.4958,
+      "step": 15804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.04198956489563,
+      "learning_rate": 4.5824742268041235e-05,
+      "loss": 1.4967,
+      "step": 15805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7698755264282227,
+      "learning_rate": 4.5824477927570715e-05,
+      "loss": 1.4672,
+      "step": 15806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6688206195831299,
+      "learning_rate": 4.582421358710019e-05,
+      "loss": 1.4002,
+      "step": 15807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.617724895477295,
+      "learning_rate": 4.582394924662966e-05,
+      "loss": 1.3562,
+      "step": 15808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2770729064941406,
+      "learning_rate": 4.5823684906159136e-05,
+      "loss": 1.4194,
+      "step": 15809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.552716612815857,
+      "learning_rate": 4.582342056568861e-05,
+      "loss": 1.3244,
+      "step": 15810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.86753511428833,
+      "learning_rate": 4.582315622521808e-05,
+      "loss": 1.3522,
+      "step": 15811
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1197891235351562,
+      "learning_rate": 4.582289188474756e-05,
+      "loss": 1.5699,
+      "step": 15812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.513650894165039,
+      "learning_rate": 4.582262754427703e-05,
+      "loss": 2.2758,
+      "step": 15813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6684037446975708,
+      "learning_rate": 4.5822363203806504e-05,
+      "loss": 1.3559,
+      "step": 15814
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6672415733337402,
+      "learning_rate": 4.582209886333598e-05,
+      "loss": 2.3777,
+      "step": 15815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8080998659133911,
+      "learning_rate": 4.582183452286545e-05,
+      "loss": 1.4626,
+      "step": 15816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5452278852462769,
+      "learning_rate": 4.582157018239493e-05,
+      "loss": 1.1544,
+      "step": 15817
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7081403732299805,
+      "learning_rate": 4.58213058419244e-05,
+      "loss": 2.0234,
+      "step": 15818
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.59856116771698,
+      "learning_rate": 4.582104150145388e-05,
+      "loss": 1.4716,
+      "step": 15819
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6712661981582642,
+      "learning_rate": 4.5820777160983345e-05,
+      "loss": 2.0849,
+      "step": 15820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4345309734344482,
+      "learning_rate": 4.5820512820512826e-05,
+      "loss": 1.3384,
+      "step": 15821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5146023035049438,
+      "learning_rate": 4.58202484800423e-05,
+      "loss": 1.2978,
+      "step": 15822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4553331136703491,
+      "learning_rate": 4.581998413957177e-05,
+      "loss": 1.4031,
+      "step": 15823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7539079189300537,
+      "learning_rate": 4.5819719799101246e-05,
+      "loss": 1.4511,
+      "step": 15824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4667989015579224,
+      "learning_rate": 4.581945545863072e-05,
+      "loss": 1.1433,
+      "step": 15825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5977216958999634,
+      "learning_rate": 4.5819191118160194e-05,
+      "loss": 1.3557,
+      "step": 15826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9930044412612915,
+      "learning_rate": 4.581892677768967e-05,
+      "loss": 1.6071,
+      "step": 15827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.022120952606201,
+      "learning_rate": 4.581866243721914e-05,
+      "loss": 1.5473,
+      "step": 15828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9263719320297241,
+      "learning_rate": 4.5818398096748614e-05,
+      "loss": 1.3316,
+      "step": 15829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7055995464324951,
+      "learning_rate": 4.581813375627809e-05,
+      "loss": 1.3802,
+      "step": 15830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.89972984790802,
+      "learning_rate": 4.581786941580756e-05,
+      "loss": 1.4813,
+      "step": 15831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7632683515548706,
+      "learning_rate": 4.581760507533704e-05,
+      "loss": 1.4779,
+      "step": 15832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8307881355285645,
+      "learning_rate": 4.581734073486651e-05,
+      "loss": 1.3835,
+      "step": 15833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.134925603866577,
+      "learning_rate": 4.581707639439599e-05,
+      "loss": 1.3595,
+      "step": 15834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.615947961807251,
+      "learning_rate": 4.5816812053925456e-05,
+      "loss": 1.5237,
+      "step": 15835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7758313417434692,
+      "learning_rate": 4.581654771345493e-05,
+      "loss": 1.4735,
+      "step": 15836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9821381568908691,
+      "learning_rate": 4.581628337298441e-05,
+      "loss": 1.3395,
+      "step": 15837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7750473022460938,
+      "learning_rate": 4.5816019032513876e-05,
+      "loss": 1.4894,
+      "step": 15838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1376700401306152,
+      "learning_rate": 4.581575469204336e-05,
+      "loss": 1.4102,
+      "step": 15839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0117921829223633,
+      "learning_rate": 4.5815490351572824e-05,
+      "loss": 1.4296,
+      "step": 15840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7695198059082031,
+      "learning_rate": 4.5815226011102304e-05,
+      "loss": 1.2767,
+      "step": 15841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9672430753707886,
+      "learning_rate": 4.581496167063178e-05,
+      "loss": 1.546,
+      "step": 15842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6303247213363647,
+      "learning_rate": 4.581469733016125e-05,
+      "loss": 1.4726,
+      "step": 15843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7848395109176636,
+      "learning_rate": 4.5814432989690725e-05,
+      "loss": 1.4853,
+      "step": 15844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8675663471221924,
+      "learning_rate": 4.58141686492202e-05,
+      "loss": 1.3768,
+      "step": 15845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.46181583404541,
+      "learning_rate": 4.581390430874967e-05,
+      "loss": 1.4748,
+      "step": 15846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9565788507461548,
+      "learning_rate": 4.5813639968279145e-05,
+      "loss": 1.3184,
+      "step": 15847
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1141092777252197,
+      "learning_rate": 4.581337562780862e-05,
+      "loss": 1.3084,
+      "step": 15848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7646136283874512,
+      "learning_rate": 4.581311128733809e-05,
+      "loss": 1.2491,
+      "step": 15849
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5653595924377441,
+      "learning_rate": 4.5812846946867566e-05,
+      "loss": 1.293,
+      "step": 15850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1672911643981934,
+      "learning_rate": 4.581258260639704e-05,
+      "loss": 1.7345,
+      "step": 15851
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.195446729660034,
+      "learning_rate": 4.581231826592652e-05,
+      "loss": 1.0976,
+      "step": 15852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4793461561203003,
+      "learning_rate": 4.581205392545599e-05,
+      "loss": 1.4792,
+      "step": 15853
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8298628330230713,
+      "learning_rate": 4.581178958498547e-05,
+      "loss": 1.4441,
+      "step": 15854
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.783158302307129,
+      "learning_rate": 4.5811525244514934e-05,
+      "loss": 2.3511,
+      "step": 15855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.872558116912842,
+      "learning_rate": 4.5811260904044414e-05,
+      "loss": 1.3891,
+      "step": 15856
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.609777808189392,
+      "learning_rate": 4.581099656357389e-05,
+      "loss": 1.6938,
+      "step": 15857
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7961498498916626,
+      "learning_rate": 4.581073222310336e-05,
+      "loss": 1.4048,
+      "step": 15858
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8432390689849854,
+      "learning_rate": 4.5810467882632835e-05,
+      "loss": 1.3561,
+      "step": 15859
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7146918773651123,
+      "learning_rate": 4.581020354216231e-05,
+      "loss": 1.7911,
+      "step": 15860
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2930190563201904,
+      "learning_rate": 4.580993920169178e-05,
+      "loss": 1.5466,
+      "step": 15861
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1291236877441406,
+      "learning_rate": 4.5809674861221256e-05,
+      "loss": 1.5332,
+      "step": 15862
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7033205032348633,
+      "learning_rate": 4.580941052075073e-05,
+      "loss": 1.4383,
+      "step": 15863
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3934571743011475,
+      "learning_rate": 4.58091461802802e-05,
+      "loss": 1.5942,
+      "step": 15864
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3943183422088623,
+      "learning_rate": 4.5808881839809676e-05,
+      "loss": 1.3728,
+      "step": 15865
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.705928921699524,
+      "learning_rate": 4.580861749933915e-05,
+      "loss": 2.0599,
+      "step": 15866
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.103001832962036,
+      "learning_rate": 4.580835315886863e-05,
+      "loss": 1.5038,
+      "step": 15867
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.884567141532898,
+      "learning_rate": 4.58080888183981e-05,
+      "loss": 1.3842,
+      "step": 15868
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4469529390335083,
+      "learning_rate": 4.580782447792758e-05,
+      "loss": 1.2734,
+      "step": 15869
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5490257740020752,
+      "learning_rate": 4.5807560137457044e-05,
+      "loss": 1.28,
+      "step": 15870
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7910321950912476,
+      "learning_rate": 4.5807295796986525e-05,
+      "loss": 1.5455,
+      "step": 15871
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.350740671157837,
+      "learning_rate": 4.5807031456516e-05,
+      "loss": 1.3218,
+      "step": 15872
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9376500844955444,
+      "learning_rate": 4.5806767116045465e-05,
+      "loss": 1.4265,
+      "step": 15873
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.893728256225586,
+      "learning_rate": 4.5806502775574945e-05,
+      "loss": 1.426,
+      "step": 15874
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8152507543563843,
+      "learning_rate": 4.580623843510441e-05,
+      "loss": 1.4272,
+      "step": 15875
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5648423433303833,
+      "learning_rate": 4.580597409463389e-05,
+      "loss": 1.3129,
+      "step": 15876
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9713670015335083,
+      "learning_rate": 4.5805709754163366e-05,
+      "loss": 1.5369,
+      "step": 15877
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7981244325637817,
+      "learning_rate": 4.580544541369284e-05,
+      "loss": 1.3865,
+      "step": 15878
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3695958852767944,
+      "learning_rate": 4.580518107322231e-05,
+      "loss": 1.2809,
+      "step": 15879
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9877967834472656,
+      "learning_rate": 4.580491673275179e-05,
+      "loss": 1.4305,
+      "step": 15880
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.063976526260376,
+      "learning_rate": 4.580465239228126e-05,
+      "loss": 1.4236,
+      "step": 15881
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9335886240005493,
+      "learning_rate": 4.5804388051810734e-05,
+      "loss": 1.3323,
+      "step": 15882
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3110241889953613,
+      "learning_rate": 4.580412371134021e-05,
+      "loss": 1.4409,
+      "step": 15883
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.759422779083252,
+      "learning_rate": 4.580385937086968e-05,
+      "loss": 1.3281,
+      "step": 15884
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.623232364654541,
+      "learning_rate": 4.5803595030399155e-05,
+      "loss": 1.6095,
+      "step": 15885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5829142332077026,
+      "learning_rate": 4.580333068992863e-05,
+      "loss": 1.2635,
+      "step": 15886
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.508341670036316,
+      "learning_rate": 4.580306634945811e-05,
+      "loss": 1.3641,
+      "step": 15887
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2013227939605713,
+      "learning_rate": 4.5802802008987575e-05,
+      "loss": 1.5309,
+      "step": 15888
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7103885412216187,
+      "learning_rate": 4.5802537668517056e-05,
+      "loss": 1.3297,
+      "step": 15889
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.868031620979309,
+      "learning_rate": 4.580227332804652e-05,
+      "loss": 1.3651,
+      "step": 15890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.879772424697876,
+      "learning_rate": 4.5802008987576e-05,
+      "loss": 1.3369,
+      "step": 15891
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3952988386154175,
+      "learning_rate": 4.5801744647105476e-05,
+      "loss": 1.4042,
+      "step": 15892
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8486242294311523,
+      "learning_rate": 4.580148030663495e-05,
+      "loss": 1.3398,
+      "step": 15893
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.715949296951294,
+      "learning_rate": 4.5801215966164424e-05,
+      "loss": 1.3426,
+      "step": 15894
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.39567232131958,
+      "learning_rate": 4.58009516256939e-05,
+      "loss": 1.3539,
+      "step": 15895
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9795478582382202,
+      "learning_rate": 4.580068728522337e-05,
+      "loss": 1.4277,
+      "step": 15896
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.442605495452881,
+      "learning_rate": 4.5800422944752844e-05,
+      "loss": 1.5604,
+      "step": 15897
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5609560012817383,
+      "learning_rate": 4.580015860428232e-05,
+      "loss": 1.2988,
+      "step": 15898
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3693859577178955,
+      "learning_rate": 4.579989426381179e-05,
+      "loss": 1.528,
+      "step": 15899
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.634560227394104,
+      "learning_rate": 4.5799629923341265e-05,
+      "loss": 1.209,
+      "step": 15900
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7379933595657349,
+      "learning_rate": 4.579936558287074e-05,
+      "loss": 1.2791,
+      "step": 15901
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6569349765777588,
+      "learning_rate": 4.579910124240022e-05,
+      "loss": 1.3486,
+      "step": 15902
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2379684448242188,
+      "learning_rate": 4.5798836901929686e-05,
+      "loss": 1.3641,
+      "step": 15903
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.896633505821228,
+      "learning_rate": 4.5798572561459166e-05,
+      "loss": 1.3628,
+      "step": 15904
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6300926208496094,
+      "learning_rate": 4.579830822098863e-05,
+      "loss": 1.4321,
+      "step": 15905
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.910274624824524,
+      "learning_rate": 4.579804388051811e-05,
+      "loss": 1.3866,
+      "step": 15906
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8560822010040283,
+      "learning_rate": 4.579777954004759e-05,
+      "loss": 1.4503,
+      "step": 15907
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7524470090866089,
+      "learning_rate": 4.5797515199577054e-05,
+      "loss": 1.2423,
+      "step": 15908
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7108914852142334,
+      "learning_rate": 4.5797250859106534e-05,
+      "loss": 1.356,
+      "step": 15909
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1939780712127686,
+      "learning_rate": 4.5796986518636e-05,
+      "loss": 1.3718,
+      "step": 15910
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4770855903625488,
+      "learning_rate": 4.579672217816548e-05,
+      "loss": 1.2776,
+      "step": 15911
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4274659156799316,
+      "learning_rate": 4.5796457837694955e-05,
+      "loss": 1.4824,
+      "step": 15912
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5380388498306274,
+      "learning_rate": 4.579619349722443e-05,
+      "loss": 1.2359,
+      "step": 15913
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9178667068481445,
+      "learning_rate": 4.57959291567539e-05,
+      "loss": 1.5094,
+      "step": 15914
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0710253715515137,
+      "learning_rate": 4.5795664816283375e-05,
+      "loss": 1.4884,
+      "step": 15915
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1708247661590576,
+      "learning_rate": 4.579540047581285e-05,
+      "loss": 1.5955,
+      "step": 15916
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9690275192260742,
+      "learning_rate": 4.579513613534233e-05,
+      "loss": 1.3817,
+      "step": 15917
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5238637924194336,
+      "learning_rate": 4.5794871794871796e-05,
+      "loss": 1.4082,
+      "step": 15918
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.900524616241455,
+      "learning_rate": 4.579460745440127e-05,
+      "loss": 1.5476,
+      "step": 15919
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.884793996810913,
+      "learning_rate": 4.579434311393074e-05,
+      "loss": 1.3337,
+      "step": 15920
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6718556880950928,
+      "learning_rate": 4.579407877346022e-05,
+      "loss": 1.3502,
+      "step": 15921
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9170821905136108,
+      "learning_rate": 4.57938144329897e-05,
+      "loss": 1.3813,
+      "step": 15922
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8224074840545654,
+      "learning_rate": 4.5793550092519164e-05,
+      "loss": 1.3788,
+      "step": 15923
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4376404285430908,
+      "learning_rate": 4.5793285752048644e-05,
+      "loss": 1.2945,
+      "step": 15924
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5655109882354736,
+      "learning_rate": 4.579302141157811e-05,
+      "loss": 1.3543,
+      "step": 15925
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6647381782531738,
+      "learning_rate": 4.579275707110759e-05,
+      "loss": 1.3859,
+      "step": 15926
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.663139820098877,
+      "learning_rate": 4.5792492730637065e-05,
+      "loss": 1.4265,
+      "step": 15927
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6630892753601074,
+      "learning_rate": 4.579222839016654e-05,
+      "loss": 1.724,
+      "step": 15928
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9421981573104858,
+      "learning_rate": 4.579196404969601e-05,
+      "loss": 1.3547,
+      "step": 15929
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5971505641937256,
+      "learning_rate": 4.5791699709225486e-05,
+      "loss": 1.3719,
+      "step": 15930
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4191902875900269,
+      "learning_rate": 4.579143536875496e-05,
+      "loss": 1.3225,
+      "step": 15931
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.223417043685913,
+      "learning_rate": 4.579117102828443e-05,
+      "loss": 1.4445,
+      "step": 15932
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1872801780700684,
+      "learning_rate": 4.5790906687813906e-05,
+      "loss": 1.4435,
+      "step": 15933
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5418782234191895,
+      "learning_rate": 4.579064234734338e-05,
+      "loss": 1.3725,
+      "step": 15934
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3938138484954834,
+      "learning_rate": 4.5790378006872853e-05,
+      "loss": 1.4728,
+      "step": 15935
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1349987983703613,
+      "learning_rate": 4.579011366640233e-05,
+      "loss": 1.3883,
+      "step": 15936
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5180929899215698,
+      "learning_rate": 4.578984932593181e-05,
+      "loss": 1.3092,
+      "step": 15937
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0316174030303955,
+      "learning_rate": 4.5789584985461274e-05,
+      "loss": 1.447,
+      "step": 15938
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.296497106552124,
+      "learning_rate": 4.5789320644990755e-05,
+      "loss": 1.0193,
+      "step": 15939
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4902359247207642,
+      "learning_rate": 4.578905630452022e-05,
+      "loss": 1.2186,
+      "step": 15940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8593178987503052,
+      "learning_rate": 4.57887919640497e-05,
+      "loss": 1.2948,
+      "step": 15941
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.146162271499634,
+      "learning_rate": 4.5788527623579175e-05,
+      "loss": 1.6062,
+      "step": 15942
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8281097412109375,
+      "learning_rate": 4.578826328310865e-05,
+      "loss": 1.4052,
+      "step": 15943
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2118802070617676,
+      "learning_rate": 4.578799894263812e-05,
+      "loss": 1.4501,
+      "step": 15944
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.671939730644226,
+      "learning_rate": 4.578773460216759e-05,
+      "loss": 1.3561,
+      "step": 15945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5182392597198486,
+      "learning_rate": 4.578747026169707e-05,
+      "loss": 1.3421,
+      "step": 15946
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9192917346954346,
+      "learning_rate": 4.578720592122654e-05,
+      "loss": 1.6561,
+      "step": 15947
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6419705152511597,
+      "learning_rate": 4.578694158075602e-05,
+      "loss": 1.4275,
+      "step": 15948
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5287798643112183,
+      "learning_rate": 4.578667724028549e-05,
+      "loss": 1.2408,
+      "step": 15949
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4941140413284302,
+      "learning_rate": 4.5786412899814964e-05,
+      "loss": 1.2257,
+      "step": 15950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7869422435760498,
+      "learning_rate": 4.578614855934444e-05,
+      "loss": 1.3754,
+      "step": 15951
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9810739755630493,
+      "learning_rate": 4.578588421887392e-05,
+      "loss": 1.5379,
+      "step": 15952
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.450097680091858,
+      "learning_rate": 4.5785619878403385e-05,
+      "loss": 1.246,
+      "step": 15953
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9983086585998535,
+      "learning_rate": 4.578535553793286e-05,
+      "loss": 1.4546,
+      "step": 15954
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.534934401512146,
+      "learning_rate": 4.578509119746233e-05,
+      "loss": 1.3583,
+      "step": 15955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8025199174880981,
+      "learning_rate": 4.5784826856991805e-05,
+      "loss": 1.2956,
+      "step": 15956
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7607831954956055,
+      "learning_rate": 4.5784562516521286e-05,
+      "loss": 1.634,
+      "step": 15957
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.114410400390625,
+      "learning_rate": 4.578429817605075e-05,
+      "loss": 1.5812,
+      "step": 15958
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.588179588317871,
+      "learning_rate": 4.578403383558023e-05,
+      "loss": 1.3128,
+      "step": 15959
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4453015327453613,
+      "learning_rate": 4.57837694951097e-05,
+      "loss": 1.4715,
+      "step": 15960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0402252674102783,
+      "learning_rate": 4.578350515463918e-05,
+      "loss": 1.3605,
+      "step": 15961
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7083618640899658,
+      "learning_rate": 4.5783240814168653e-05,
+      "loss": 1.2094,
+      "step": 15962
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8187408447265625,
+      "learning_rate": 4.578297647369813e-05,
+      "loss": 1.5129,
+      "step": 15963
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.811582088470459,
+      "learning_rate": 4.57827121332276e-05,
+      "loss": 1.6047,
+      "step": 15964
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6296755075454712,
+      "learning_rate": 4.5782447792757074e-05,
+      "loss": 1.2481,
+      "step": 15965
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5322773456573486,
+      "learning_rate": 4.578218345228655e-05,
+      "loss": 1.9263,
+      "step": 15966
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0140559673309326,
+      "learning_rate": 4.578191911181602e-05,
+      "loss": 1.3939,
+      "step": 15967
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.385033369064331,
+      "learning_rate": 4.5781654771345495e-05,
+      "loss": 1.2798,
+      "step": 15968
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8482698202133179,
+      "learning_rate": 4.578139043087497e-05,
+      "loss": 1.3212,
+      "step": 15969
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5677636861801147,
+      "learning_rate": 4.578112609040444e-05,
+      "loss": 2.2302,
+      "step": 15970
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.654048204421997,
+      "learning_rate": 4.5780861749933916e-05,
+      "loss": 1.9785,
+      "step": 15971
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2433969974517822,
+      "learning_rate": 4.578059740946339e-05,
+      "loss": 1.2402,
+      "step": 15972
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2909021377563477,
+      "learning_rate": 4.578033306899286e-05,
+      "loss": 1.5729,
+      "step": 15973
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1778790950775146,
+      "learning_rate": 4.578006872852234e-05,
+      "loss": 1.5225,
+      "step": 15974
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4783360958099365,
+      "learning_rate": 4.577980438805181e-05,
+      "loss": 1.4272,
+      "step": 15975
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0973780155181885,
+      "learning_rate": 4.577954004758129e-05,
+      "loss": 1.4224,
+      "step": 15976
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9192630052566528,
+      "learning_rate": 4.577927570711076e-05,
+      "loss": 1.2352,
+      "step": 15977
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.845144033432007,
+      "learning_rate": 4.577901136664024e-05,
+      "loss": 1.4529,
+      "step": 15978
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6937774419784546,
+      "learning_rate": 4.577874702616971e-05,
+      "loss": 1.266,
+      "step": 15979
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5753854513168335,
+      "learning_rate": 4.577848268569918e-05,
+      "loss": 1.2698,
+      "step": 15980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9873030185699463,
+      "learning_rate": 4.577821834522866e-05,
+      "loss": 1.4442,
+      "step": 15981
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1007566452026367,
+      "learning_rate": 4.5777954004758125e-05,
+      "loss": 1.403,
+      "step": 15982
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.427070140838623,
+      "learning_rate": 4.5777689664287605e-05,
+      "loss": 1.1883,
+      "step": 15983
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8961056470870972,
+      "learning_rate": 4.577742532381708e-05,
+      "loss": 1.5091,
+      "step": 15984
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3537838459014893,
+      "learning_rate": 4.577716098334655e-05,
+      "loss": 1.3634,
+      "step": 15985
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9435783624649048,
+      "learning_rate": 4.5776896642876026e-05,
+      "loss": 1.5052,
+      "step": 15986
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7879990339279175,
+      "learning_rate": 4.57766323024055e-05,
+      "loss": 1.3118,
+      "step": 15987
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0320017337799072,
+      "learning_rate": 4.577636796193497e-05,
+      "loss": 1.6119,
+      "step": 15988
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.69572377204895,
+      "learning_rate": 4.5776103621464453e-05,
+      "loss": 1.6656,
+      "step": 15989
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6772574186325073,
+      "learning_rate": 4.577583928099392e-05,
+      "loss": 1.4561,
+      "step": 15990
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8247363567352295,
+      "learning_rate": 4.5775574940523394e-05,
+      "loss": 1.4199,
+      "step": 15991
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6004376411437988,
+      "learning_rate": 4.577531060005287e-05,
+      "loss": 1.2681,
+      "step": 15992
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5169178247451782,
+      "learning_rate": 4.577504625958234e-05,
+      "loss": 1.286,
+      "step": 15993
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.055220365524292,
+      "learning_rate": 4.577478191911182e-05,
+      "loss": 1.4379,
+      "step": 15994
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.054744243621826,
+      "learning_rate": 4.577451757864129e-05,
+      "loss": 1.3224,
+      "step": 15995
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9708679914474487,
+      "learning_rate": 4.577425323817077e-05,
+      "loss": 1.5226,
+      "step": 15996
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.687187671661377,
+      "learning_rate": 4.5773988897700235e-05,
+      "loss": 2.2181,
+      "step": 15997
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8306703567504883,
+      "learning_rate": 4.5773724557229716e-05,
+      "loss": 1.448,
+      "step": 15998
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6633774042129517,
+      "learning_rate": 4.577346021675919e-05,
+      "loss": 1.9544,
+      "step": 15999
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.112058639526367,
+      "learning_rate": 4.577319587628866e-05,
+      "loss": 1.555,
+      "step": 16000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.142777442932129,
+      "learning_rate": 4.5772931535818136e-05,
+      "loss": 1.4962,
+      "step": 16001
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.822399616241455,
+      "learning_rate": 4.577266719534761e-05,
+      "loss": 1.2991,
+      "step": 16002
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9395054578781128,
+      "learning_rate": 4.5772402854877083e-05,
+      "loss": 1.3794,
+      "step": 16003
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.147106885910034,
+      "learning_rate": 4.577213851440656e-05,
+      "loss": 1.6524,
+      "step": 16004
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7428292036056519,
+      "learning_rate": 4.577187417393603e-05,
+      "loss": 1.2864,
+      "step": 16005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5940158367156982,
+      "learning_rate": 4.5771609833465504e-05,
+      "loss": 1.2994,
+      "step": 16006
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.378143310546875,
+      "learning_rate": 4.577134549299498e-05,
+      "loss": 1.4114,
+      "step": 16007
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7180525064468384,
+      "learning_rate": 4.577108115252445e-05,
+      "loss": 1.2821,
+      "step": 16008
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6578139066696167,
+      "learning_rate": 4.577081681205393e-05,
+      "loss": 1.3506,
+      "step": 16009
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8440682888031006,
+      "learning_rate": 4.57705524715834e-05,
+      "loss": 1.3902,
+      "step": 16010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.844650149345398,
+      "learning_rate": 4.577028813111288e-05,
+      "loss": 1.3171,
+      "step": 16011
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4940836429595947,
+      "learning_rate": 4.5770023790642346e-05,
+      "loss": 1.5381,
+      "step": 16012
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7137423753738403,
+      "learning_rate": 4.5769759450171826e-05,
+      "loss": 1.812,
+      "step": 16013
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7592445611953735,
+      "learning_rate": 4.57694951097013e-05,
+      "loss": 1.3516,
+      "step": 16014
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8646883964538574,
+      "learning_rate": 4.576923076923077e-05,
+      "loss": 1.3742,
+      "step": 16015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5813971757888794,
+      "learning_rate": 4.5768966428760247e-05,
+      "loss": 1.4418,
+      "step": 16016
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.923256516456604,
+      "learning_rate": 4.5768702088289713e-05,
+      "loss": 1.4713,
+      "step": 16017
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5983679294586182,
+      "learning_rate": 4.5768437747819194e-05,
+      "loss": 1.2955,
+      "step": 16018
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7157742977142334,
+      "learning_rate": 4.576817340734867e-05,
+      "loss": 1.3949,
+      "step": 16019
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5591239929199219,
+      "learning_rate": 4.576790906687814e-05,
+      "loss": 1.3877,
+      "step": 16020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9998621940612793,
+      "learning_rate": 4.5767644726407614e-05,
+      "loss": 1.3579,
+      "step": 16021
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.734366536140442,
+      "learning_rate": 4.576738038593709e-05,
+      "loss": 1.3127,
+      "step": 16022
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5637400150299072,
+      "learning_rate": 4.576711604546656e-05,
+      "loss": 1.219,
+      "step": 16023
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.176999092102051,
+      "learning_rate": 4.576685170499604e-05,
+      "loss": 1.3794,
+      "step": 16024
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.176679849624634,
+      "learning_rate": 4.576658736452551e-05,
+      "loss": 1.4282,
+      "step": 16025
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8916112184524536,
+      "learning_rate": 4.576632302405498e-05,
+      "loss": 1.4614,
+      "step": 16026
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8270659446716309,
+      "learning_rate": 4.5766058683584456e-05,
+      "loss": 1.5091,
+      "step": 16027
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5426018238067627,
+      "learning_rate": 4.576579434311393e-05,
+      "loss": 1.2295,
+      "step": 16028
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1575708389282227,
+      "learning_rate": 4.576553000264341e-05,
+      "loss": 1.4201,
+      "step": 16029
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.687066674232483,
+      "learning_rate": 4.5765265662172877e-05,
+      "loss": 1.3778,
+      "step": 16030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.045137405395508,
+      "learning_rate": 4.576500132170236e-05,
+      "loss": 1.4398,
+      "step": 16031
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5042141675949097,
+      "learning_rate": 4.5764736981231824e-05,
+      "loss": 1.286,
+      "step": 16032
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8844164609909058,
+      "learning_rate": 4.5764472640761304e-05,
+      "loss": 1.3573,
+      "step": 16033
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.735732078552246,
+      "learning_rate": 4.576420830029078e-05,
+      "loss": 1.229,
+      "step": 16034
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7780835628509521,
+      "learning_rate": 4.576394395982025e-05,
+      "loss": 1.4023,
+      "step": 16035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.526819109916687,
+      "learning_rate": 4.5763679619349725e-05,
+      "loss": 1.3692,
+      "step": 16036
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8225048780441284,
+      "learning_rate": 4.57634152788792e-05,
+      "loss": 1.1974,
+      "step": 16037
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5332499742507935,
+      "learning_rate": 4.576315093840867e-05,
+      "loss": 1.2593,
+      "step": 16038
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.870290517807007,
+      "learning_rate": 4.5762886597938146e-05,
+      "loss": 1.3718,
+      "step": 16039
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5942630767822266,
+      "learning_rate": 4.576262225746762e-05,
+      "loss": 1.423,
+      "step": 16040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9006657600402832,
+      "learning_rate": 4.576235791699709e-05,
+      "loss": 1.5095,
+      "step": 16041
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5669147968292236,
+      "learning_rate": 4.5762093576526566e-05,
+      "loss": 1.3283,
+      "step": 16042
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8985673189163208,
+      "learning_rate": 4.576182923605604e-05,
+      "loss": 1.5554,
+      "step": 16043
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7031934261322021,
+      "learning_rate": 4.576156489558552e-05,
+      "loss": 1.4028,
+      "step": 16044
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8661324977874756,
+      "learning_rate": 4.576130055511499e-05,
+      "loss": 1.3715,
+      "step": 16045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6700475215911865,
+      "learning_rate": 4.576103621464447e-05,
+      "loss": 1.4642,
+      "step": 16046
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4814807176589966,
+      "learning_rate": 4.5760771874173934e-05,
+      "loss": 1.3292,
+      "step": 16047
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.644482970237732,
+      "learning_rate": 4.5760507533703414e-05,
+      "loss": 1.3228,
+      "step": 16048
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4879400730133057,
+      "learning_rate": 4.576024319323289e-05,
+      "loss": 1.4201,
+      "step": 16049
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3160879611968994,
+      "learning_rate": 4.575997885276236e-05,
+      "loss": 1.4516,
+      "step": 16050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.650795817375183,
+      "learning_rate": 4.5759714512291835e-05,
+      "loss": 1.4062,
+      "step": 16051
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6750842332839966,
+      "learning_rate": 4.575945017182131e-05,
+      "loss": 1.4297,
+      "step": 16052
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.028087615966797,
+      "learning_rate": 4.575918583135078e-05,
+      "loss": 1.473,
+      "step": 16053
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8154969215393066,
+      "learning_rate": 4.5758921490880256e-05,
+      "loss": 1.2179,
+      "step": 16054
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5908477306365967,
+      "learning_rate": 4.575865715040973e-05,
+      "loss": 1.5363,
+      "step": 16055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0666182041168213,
+      "learning_rate": 4.57583928099392e-05,
+      "loss": 1.4156,
+      "step": 16056
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6964613199234009,
+      "learning_rate": 4.5758128469468677e-05,
+      "loss": 1.3182,
+      "step": 16057
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6936659812927246,
+      "learning_rate": 4.575786412899815e-05,
+      "loss": 1.4557,
+      "step": 16058
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5630062818527222,
+      "learning_rate": 4.575759978852763e-05,
+      "loss": 1.4269,
+      "step": 16059
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0621302127838135,
+      "learning_rate": 4.57573354480571e-05,
+      "loss": 1.4201,
+      "step": 16060
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8545985221862793,
+      "learning_rate": 4.575707110758658e-05,
+      "loss": 1.3516,
+      "step": 16061
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8183033466339111,
+      "learning_rate": 4.5756806767116044e-05,
+      "loss": 1.3382,
+      "step": 16062
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.857954740524292,
+      "learning_rate": 4.575654242664552e-05,
+      "loss": 1.4011,
+      "step": 16063
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8144150972366333,
+      "learning_rate": 4.5756278086175e-05,
+      "loss": 1.462,
+      "step": 16064
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.476051092147827,
+      "learning_rate": 4.5756013745704465e-05,
+      "loss": 1.384,
+      "step": 16065
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.975104570388794,
+      "learning_rate": 4.5755749405233945e-05,
+      "loss": 1.447,
+      "step": 16066
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7864620685577393,
+      "learning_rate": 4.575548506476341e-05,
+      "loss": 1.3865,
+      "step": 16067
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.751600742340088,
+      "learning_rate": 4.575522072429289e-05,
+      "loss": 1.2939,
+      "step": 16068
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8567336797714233,
+      "learning_rate": 4.5754956383822366e-05,
+      "loss": 1.3168,
+      "step": 16069
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0211424827575684,
+      "learning_rate": 4.575469204335184e-05,
+      "loss": 1.5681,
+      "step": 16070
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9317595958709717,
+      "learning_rate": 4.575442770288131e-05,
+      "loss": 1.3944,
+      "step": 16071
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6798359155654907,
+      "learning_rate": 4.575416336241079e-05,
+      "loss": 1.3166,
+      "step": 16072
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1849536895751953,
+      "learning_rate": 4.575389902194026e-05,
+      "loss": 1.3759,
+      "step": 16073
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6953059434890747,
+      "learning_rate": 4.5753634681469734e-05,
+      "loss": 1.2073,
+      "step": 16074
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8477764129638672,
+      "learning_rate": 4.575337034099921e-05,
+      "loss": 1.3889,
+      "step": 16075
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.429535984992981,
+      "learning_rate": 4.575310600052868e-05,
+      "loss": 1.2409,
+      "step": 16076
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6133451461791992,
+      "learning_rate": 4.5752841660058155e-05,
+      "loss": 1.3498,
+      "step": 16077
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4192509651184082,
+      "learning_rate": 4.575257731958763e-05,
+      "loss": 1.3111,
+      "step": 16078
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6072689294815063,
+      "learning_rate": 4.575231297911711e-05,
+      "loss": 1.4291,
+      "step": 16079
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.495136260986328,
+      "learning_rate": 4.5752048638646575e-05,
+      "loss": 1.5082,
+      "step": 16080
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6790111064910889,
+      "learning_rate": 4.5751784298176056e-05,
+      "loss": 1.5018,
+      "step": 16081
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6367504596710205,
+      "learning_rate": 4.575151995770552e-05,
+      "loss": 1.3377,
+      "step": 16082
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9058043956756592,
+      "learning_rate": 4.5751255617235e-05,
+      "loss": 1.3845,
+      "step": 16083
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.441527843475342,
+      "learning_rate": 4.5750991276764477e-05,
+      "loss": 1.4116,
+      "step": 16084
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3805280923843384,
+      "learning_rate": 4.575072693629395e-05,
+      "loss": 1.3082,
+      "step": 16085
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.628935694694519,
+      "learning_rate": 4.5750462595823424e-05,
+      "loss": 1.4036,
+      "step": 16086
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6973533630371094,
+      "learning_rate": 4.57501982553529e-05,
+      "loss": 1.3605,
+      "step": 16087
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.312047004699707,
+      "learning_rate": 4.574993391488237e-05,
+      "loss": 1.2338,
+      "step": 16088
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2257750034332275,
+      "learning_rate": 4.5749669574411844e-05,
+      "loss": 1.4685,
+      "step": 16089
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7400206327438354,
+      "learning_rate": 4.574940523394132e-05,
+      "loss": 1.4021,
+      "step": 16090
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3862240314483643,
+      "learning_rate": 4.574914089347079e-05,
+      "loss": 1.4288,
+      "step": 16091
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4760903120040894,
+      "learning_rate": 4.5748876553000265e-05,
+      "loss": 2.2579,
+      "step": 16092
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5195598602294922,
+      "learning_rate": 4.574861221252974e-05,
+      "loss": 1.3099,
+      "step": 16093
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.594611406326294,
+      "learning_rate": 4.574834787205922e-05,
+      "loss": 1.3438,
+      "step": 16094
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0549566745758057,
+      "learning_rate": 4.5748083531588686e-05,
+      "loss": 1.3851,
+      "step": 16095
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7065590620040894,
+      "learning_rate": 4.5747819191118166e-05,
+      "loss": 1.1546,
+      "step": 16096
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7860801219940186,
+      "learning_rate": 4.574755485064763e-05,
+      "loss": 1.4565,
+      "step": 16097
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.830567479133606,
+      "learning_rate": 4.5747290510177107e-05,
+      "loss": 1.4032,
+      "step": 16098
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.591813087463379,
+      "learning_rate": 4.574702616970659e-05,
+      "loss": 1.3809,
+      "step": 16099
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5702755451202393,
+      "learning_rate": 4.5746761829236054e-05,
+      "loss": 2.2131,
+      "step": 16100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4366650581359863,
+      "learning_rate": 4.5746497488765534e-05,
+      "loss": 1.2315,
+      "step": 16101
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8054085969924927,
+      "learning_rate": 4.5746233148295e-05,
+      "loss": 1.496,
+      "step": 16102
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.165940999984741,
+      "learning_rate": 4.574596880782448e-05,
+      "loss": 1.5664,
+      "step": 16103
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6904025077819824,
+      "learning_rate": 4.5745704467353955e-05,
+      "loss": 1.4202,
+      "step": 16104
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.186940908432007,
+      "learning_rate": 4.574544012688343e-05,
+      "loss": 1.7199,
+      "step": 16105
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4593700170516968,
+      "learning_rate": 4.57451757864129e-05,
+      "loss": 1.3991,
+      "step": 16106
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.165727138519287,
+      "learning_rate": 4.5744911445942375e-05,
+      "loss": 1.6003,
+      "step": 16107
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6009248495101929,
+      "learning_rate": 4.574464710547185e-05,
+      "loss": 1.3551,
+      "step": 16108
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4047646522521973,
+      "learning_rate": 4.574438276500132e-05,
+      "loss": 1.4372,
+      "step": 16109
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2686266899108887,
+      "learning_rate": 4.5744118424530796e-05,
+      "loss": 1.3992,
+      "step": 16110
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9663071632385254,
+      "learning_rate": 4.574385408406027e-05,
+      "loss": 2.4258,
+      "step": 16111
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6165493726730347,
+      "learning_rate": 4.574358974358974e-05,
+      "loss": 1.3569,
+      "step": 16112
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.578948974609375,
+      "learning_rate": 4.574332540311922e-05,
+      "loss": 1.4135,
+      "step": 16113
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6714195013046265,
+      "learning_rate": 4.57430610626487e-05,
+      "loss": 1.4744,
+      "step": 16114
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.835322141647339,
+      "learning_rate": 4.5742796722178164e-05,
+      "loss": 1.397,
+      "step": 16115
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5655406713485718,
+      "learning_rate": 4.5742532381707644e-05,
+      "loss": 1.2078,
+      "step": 16116
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6550883054733276,
+      "learning_rate": 4.574226804123711e-05,
+      "loss": 1.457,
+      "step": 16117
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.66015362739563,
+      "learning_rate": 4.574200370076659e-05,
+      "loss": 1.4658,
+      "step": 16118
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6516317129135132,
+      "learning_rate": 4.5741739360296065e-05,
+      "loss": 1.4289,
+      "step": 16119
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.820380687713623,
+      "learning_rate": 4.574147501982554e-05,
+      "loss": 1.4929,
+      "step": 16120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4092150926589966,
+      "learning_rate": 4.574121067935501e-05,
+      "loss": 1.3369,
+      "step": 16121
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2731354236602783,
+      "learning_rate": 4.5740946338884486e-05,
+      "loss": 1.4977,
+      "step": 16122
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4983938932418823,
+      "learning_rate": 4.574068199841396e-05,
+      "loss": 1.8535,
+      "step": 16123
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0830307006835938,
+      "learning_rate": 4.574041765794343e-05,
+      "loss": 1.4178,
+      "step": 16124
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8890026807785034,
+      "learning_rate": 4.5740153317472906e-05,
+      "loss": 2.1378,
+      "step": 16125
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.587627649307251,
+      "learning_rate": 4.573988897700238e-05,
+      "loss": 1.4526,
+      "step": 16126
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.023247003555298,
+      "learning_rate": 4.5739624636531854e-05,
+      "loss": 1.4691,
+      "step": 16127
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9020836353302002,
+      "learning_rate": 4.573936029606133e-05,
+      "loss": 1.4811,
+      "step": 16128
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8059632778167725,
+      "learning_rate": 4.573909595559081e-05,
+      "loss": 1.3825,
+      "step": 16129
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7999414205551147,
+      "learning_rate": 4.5738831615120274e-05,
+      "loss": 1.4572,
+      "step": 16130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5865669250488281,
+      "learning_rate": 4.5738567274649755e-05,
+      "loss": 1.3492,
+      "step": 16131
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.338301658630371,
+      "learning_rate": 4.573830293417922e-05,
+      "loss": 1.4888,
+      "step": 16132
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.338451862335205,
+      "learning_rate": 4.57380385937087e-05,
+      "loss": 1.5901,
+      "step": 16133
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.140822649002075,
+      "learning_rate": 4.5737774253238175e-05,
+      "loss": 1.3435,
+      "step": 16134
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1193995475769043,
+      "learning_rate": 4.573750991276764e-05,
+      "loss": 1.4369,
+      "step": 16135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.022674798965454,
+      "learning_rate": 4.573724557229712e-05,
+      "loss": 1.4848,
+      "step": 16136
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.244584560394287,
+      "learning_rate": 4.573698123182659e-05,
+      "loss": 1.3194,
+      "step": 16137
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5041348934173584,
+      "learning_rate": 4.573671689135607e-05,
+      "loss": 1.4619,
+      "step": 16138
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2179372310638428,
+      "learning_rate": 4.573645255088554e-05,
+      "loss": 1.3742,
+      "step": 16139
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9104211330413818,
+      "learning_rate": 4.573618821041502e-05,
+      "loss": 1.4328,
+      "step": 16140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3894574642181396,
+      "learning_rate": 4.573592386994449e-05,
+      "loss": 1.3905,
+      "step": 16141
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.003465175628662,
+      "learning_rate": 4.5735659529473964e-05,
+      "loss": 1.3227,
+      "step": 16142
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6178185939788818,
+      "learning_rate": 4.573539518900344e-05,
+      "loss": 1.3327,
+      "step": 16143
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5766667127609253,
+      "learning_rate": 4.573513084853291e-05,
+      "loss": 1.2562,
+      "step": 16144
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5248734951019287,
+      "learning_rate": 4.5734866508062385e-05,
+      "loss": 1.3235,
+      "step": 16145
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6805185079574585,
+      "learning_rate": 4.573460216759186e-05,
+      "loss": 1.3163,
+      "step": 16146
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2135202884674072,
+      "learning_rate": 4.573433782712133e-05,
+      "loss": 1.4008,
+      "step": 16147
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5018779039382935,
+      "learning_rate": 4.5734073486650805e-05,
+      "loss": 2.0537,
+      "step": 16148
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.284942388534546,
+      "learning_rate": 4.5733809146180286e-05,
+      "loss": 1.5924,
+      "step": 16149
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.766304612159729,
+      "learning_rate": 4.573354480570975e-05,
+      "loss": 1.2914,
+      "step": 16150
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8757630586624146,
+      "learning_rate": 4.573328046523923e-05,
+      "loss": 1.4212,
+      "step": 16151
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.534593105316162,
+      "learning_rate": 4.57330161247687e-05,
+      "loss": 1.4135,
+      "step": 16152
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7853457927703857,
+      "learning_rate": 4.573275178429818e-05,
+      "loss": 1.514,
+      "step": 16153
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1593339443206787,
+      "learning_rate": 4.5732487443827654e-05,
+      "loss": 1.5852,
+      "step": 16154
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.451576828956604,
+      "learning_rate": 4.573222310335713e-05,
+      "loss": 1.3179,
+      "step": 16155
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5478413105010986,
+      "learning_rate": 4.57319587628866e-05,
+      "loss": 1.4864,
+      "step": 16156
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.299405097961426,
+      "learning_rate": 4.5731694422416074e-05,
+      "loss": 1.6353,
+      "step": 16157
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6345397233963013,
+      "learning_rate": 4.573143008194555e-05,
+      "loss": 1.3741,
+      "step": 16158
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.632604718208313,
+      "learning_rate": 4.573116574147502e-05,
+      "loss": 1.7481,
+      "step": 16159
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9876112937927246,
+      "learning_rate": 4.5730901401004495e-05,
+      "loss": 1.5255,
+      "step": 16160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4407986402511597,
+      "learning_rate": 4.573063706053397e-05,
+      "loss": 2.2557,
+      "step": 16161
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4450005292892456,
+      "learning_rate": 4.573037272006344e-05,
+      "loss": 1.3567,
+      "step": 16162
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6250503063201904,
+      "learning_rate": 4.5730108379592916e-05,
+      "loss": 1.4969,
+      "step": 16163
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8456329107284546,
+      "learning_rate": 4.5729844039122396e-05,
+      "loss": 1.3482,
+      "step": 16164
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.10274600982666,
+      "learning_rate": 4.572957969865186e-05,
+      "loss": 1.4953,
+      "step": 16165
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9596933126449585,
+      "learning_rate": 4.572931535818134e-05,
+      "loss": 1.5714,
+      "step": 16166
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.523044466972351,
+      "learning_rate": 4.572905101771081e-05,
+      "loss": 1.2224,
+      "step": 16167
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0642318725585938,
+      "learning_rate": 4.572878667724029e-05,
+      "loss": 1.3449,
+      "step": 16168
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8505996465682983,
+      "learning_rate": 4.5728522336769764e-05,
+      "loss": 1.4354,
+      "step": 16169
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.699620485305786,
+      "learning_rate": 4.572825799629923e-05,
+      "loss": 1.5806,
+      "step": 16170
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.787904143333435,
+      "learning_rate": 4.572799365582871e-05,
+      "loss": 1.2905,
+      "step": 16171
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4109954833984375,
+      "learning_rate": 4.572772931535818e-05,
+      "loss": 1.46,
+      "step": 16172
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8146921396255493,
+      "learning_rate": 4.572746497488766e-05,
+      "loss": 1.4748,
+      "step": 16173
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.102491855621338,
+      "learning_rate": 4.572720063441713e-05,
+      "loss": 1.4799,
+      "step": 16174
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1177101135253906,
+      "learning_rate": 4.5726936293946605e-05,
+      "loss": 1.3841,
+      "step": 16175
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6760021448135376,
+      "learning_rate": 4.572667195347608e-05,
+      "loss": 1.3003,
+      "step": 16176
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6231322288513184,
+      "learning_rate": 4.572640761300555e-05,
+      "loss": 1.423,
+      "step": 16177
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8602625131607056,
+      "learning_rate": 4.5726143272535026e-05,
+      "loss": 1.3141,
+      "step": 16178
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9553061723709106,
+      "learning_rate": 4.5725878932064506e-05,
+      "loss": 1.4992,
+      "step": 16179
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9031838178634644,
+      "learning_rate": 4.572561459159397e-05,
+      "loss": 1.563,
+      "step": 16180
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6737425327301025,
+      "learning_rate": 4.572535025112345e-05,
+      "loss": 1.3899,
+      "step": 16181
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5459846258163452,
+      "learning_rate": 4.572508591065292e-05,
+      "loss": 1.3016,
+      "step": 16182
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5973668098449707,
+      "learning_rate": 4.5724821570182394e-05,
+      "loss": 1.2833,
+      "step": 16183
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9337166547775269,
+      "learning_rate": 4.5724557229711874e-05,
+      "loss": 1.4612,
+      "step": 16184
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9405547380447388,
+      "learning_rate": 4.572429288924134e-05,
+      "loss": 1.3008,
+      "step": 16185
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.283811092376709,
+      "learning_rate": 4.572402854877082e-05,
+      "loss": 1.6109,
+      "step": 16186
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.1418657302856445,
+      "learning_rate": 4.572376420830029e-05,
+      "loss": 1.698,
+      "step": 16187
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.688265323638916,
+      "learning_rate": 4.572349986782977e-05,
+      "loss": 1.2425,
+      "step": 16188
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1669883728027344,
+      "learning_rate": 4.572323552735924e-05,
+      "loss": 1.416,
+      "step": 16189
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.547837257385254,
+      "learning_rate": 4.5722971186888716e-05,
+      "loss": 1.3393,
+      "step": 16190
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6857001781463623,
+      "learning_rate": 4.572270684641819e-05,
+      "loss": 1.3871,
+      "step": 16191
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8439316749572754,
+      "learning_rate": 4.572244250594766e-05,
+      "loss": 1.6629,
+      "step": 16192
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.448516845703125,
+      "learning_rate": 4.5722178165477136e-05,
+      "loss": 1.3867,
+      "step": 16193
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7396436929702759,
+      "learning_rate": 4.572191382500661e-05,
+      "loss": 1.302,
+      "step": 16194
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5985240936279297,
+      "learning_rate": 4.5721649484536084e-05,
+      "loss": 2.1238,
+      "step": 16195
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.699839472770691,
+      "learning_rate": 4.572138514406556e-05,
+      "loss": 1.4559,
+      "step": 16196
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.000148057937622,
+      "learning_rate": 4.572112080359503e-05,
+      "loss": 1.3181,
+      "step": 16197
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8533271551132202,
+      "learning_rate": 4.5720856463124504e-05,
+      "loss": 1.4774,
+      "step": 16198
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4894063472747803,
+      "learning_rate": 4.5720592122653985e-05,
+      "loss": 1.2111,
+      "step": 16199
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.867047905921936,
+      "learning_rate": 4.572032778218345e-05,
+      "loss": 1.3939,
+      "step": 16200
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8630340099334717,
+      "learning_rate": 4.572006344171293e-05,
+      "loss": 1.3724,
+      "step": 16201
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0999011993408203,
+      "learning_rate": 4.57197991012424e-05,
+      "loss": 1.4466,
+      "step": 16202
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5955922603607178,
+      "learning_rate": 4.571953476077188e-05,
+      "loss": 1.3968,
+      "step": 16203
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7215702533721924,
+      "learning_rate": 4.571927042030135e-05,
+      "loss": 1.4419,
+      "step": 16204
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.084998846054077,
+      "learning_rate": 4.5719006079830826e-05,
+      "loss": 1.4199,
+      "step": 16205
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5883758068084717,
+      "learning_rate": 4.57187417393603e-05,
+      "loss": 1.448,
+      "step": 16206
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.013739824295044,
+      "learning_rate": 4.5718477398889766e-05,
+      "loss": 1.4377,
+      "step": 16207
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6607162952423096,
+      "learning_rate": 4.571821305841925e-05,
+      "loss": 1.3925,
+      "step": 16208
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5954265594482422,
+      "learning_rate": 4.571794871794872e-05,
+      "loss": 1.3771,
+      "step": 16209
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4132182598114014,
+      "learning_rate": 4.5717684377478194e-05,
+      "loss": 1.3021,
+      "step": 16210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5350749492645264,
+      "learning_rate": 4.571742003700767e-05,
+      "loss": 1.4902,
+      "step": 16211
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4775004386901855,
+      "learning_rate": 4.571715569653714e-05,
+      "loss": 1.2948,
+      "step": 16212
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2668991088867188,
+      "learning_rate": 4.5716891356066615e-05,
+      "loss": 1.5154,
+      "step": 16213
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9156262874603271,
+      "learning_rate": 4.5716627015596095e-05,
+      "loss": 1.3929,
+      "step": 16214
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6443687677383423,
+      "learning_rate": 4.571636267512556e-05,
+      "loss": 1.6626,
+      "step": 16215
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.718437910079956,
+      "learning_rate": 4.5716098334655035e-05,
+      "loss": 1.2868,
+      "step": 16216
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7002320289611816,
+      "learning_rate": 4.571583399418451e-05,
+      "loss": 1.3184,
+      "step": 16217
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6322518587112427,
+      "learning_rate": 4.571556965371398e-05,
+      "loss": 2.1204,
+      "step": 16218
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.907832145690918,
+      "learning_rate": 4.571530531324346e-05,
+      "loss": 1.4266,
+      "step": 16219
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5697780847549438,
+      "learning_rate": 4.571504097277293e-05,
+      "loss": 1.3077,
+      "step": 16220
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.252913236618042,
+      "learning_rate": 4.571477663230241e-05,
+      "loss": 1.4449,
+      "step": 16221
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3369245529174805,
+      "learning_rate": 4.571451229183188e-05,
+      "loss": 1.2779,
+      "step": 16222
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6453255414962769,
+      "learning_rate": 4.571424795136136e-05,
+      "loss": 1.3324,
+      "step": 16223
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6947028636932373,
+      "learning_rate": 4.571398361089083e-05,
+      "loss": 1.3576,
+      "step": 16224
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4211622476577759,
+      "learning_rate": 4.5713719270420304e-05,
+      "loss": 1.4099,
+      "step": 16225
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0750319957733154,
+      "learning_rate": 4.571345492994978e-05,
+      "loss": 1.4921,
+      "step": 16226
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8767564296722412,
+      "learning_rate": 4.571319058947925e-05,
+      "loss": 1.4965,
+      "step": 16227
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8744906187057495,
+      "learning_rate": 4.5712926249008725e-05,
+      "loss": 1.3919,
+      "step": 16228
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7326871156692505,
+      "learning_rate": 4.57126619085382e-05,
+      "loss": 1.4219,
+      "step": 16229
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.483811140060425,
+      "learning_rate": 4.571239756806767e-05,
+      "loss": 1.4304,
+      "step": 16230
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5235260725021362,
+      "learning_rate": 4.5712133227597146e-05,
+      "loss": 1.2007,
+      "step": 16231
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9530611038208008,
+      "learning_rate": 4.571186888712662e-05,
+      "loss": 1.4781,
+      "step": 16232
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7517495155334473,
+      "learning_rate": 4.571160454665609e-05,
+      "loss": 1.3578,
+      "step": 16233
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5995862483978271,
+      "learning_rate": 4.571134020618557e-05,
+      "loss": 1.2757,
+      "step": 16234
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0371310710906982,
+      "learning_rate": 4.571107586571504e-05,
+      "loss": 1.4678,
+      "step": 16235
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9151372909545898,
+      "learning_rate": 4.571081152524452e-05,
+      "loss": 1.4079,
+      "step": 16236
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9557889699935913,
+      "learning_rate": 4.571054718477399e-05,
+      "loss": 1.4567,
+      "step": 16237
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.893230438232422,
+      "learning_rate": 4.571028284430347e-05,
+      "loss": 2.3355,
+      "step": 16238
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5564494132995605,
+      "learning_rate": 4.571001850383294e-05,
+      "loss": 1.358,
+      "step": 16239
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.220459222793579,
+      "learning_rate": 4.5709754163362415e-05,
+      "loss": 1.4934,
+      "step": 16240
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.502685785293579,
+      "learning_rate": 4.570948982289189e-05,
+      "loss": 1.2995,
+      "step": 16241
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5722193717956543,
+      "learning_rate": 4.570922548242136e-05,
+      "loss": 1.4166,
+      "step": 16242
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6150003671646118,
+      "learning_rate": 4.5708961141950835e-05,
+      "loss": 1.2788,
+      "step": 16243
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5368775129318237,
+      "learning_rate": 4.570869680148031e-05,
+      "loss": 1.1997,
+      "step": 16244
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7358211278915405,
+      "learning_rate": 4.570843246100978e-05,
+      "loss": 1.2609,
+      "step": 16245
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8768924474716187,
+      "learning_rate": 4.5708168120539256e-05,
+      "loss": 1.3665,
+      "step": 16246
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.86603844165802,
+      "learning_rate": 4.570790378006873e-05,
+      "loss": 1.4302,
+      "step": 16247
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9777591228485107,
+      "learning_rate": 4.57076394395982e-05,
+      "loss": 1.4438,
+      "step": 16248
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5504873991012573,
+      "learning_rate": 4.5707375099127684e-05,
+      "loss": 1.2711,
+      "step": 16249
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7489149570465088,
+      "learning_rate": 4.570711075865715e-05,
+      "loss": 1.4005,
+      "step": 16250
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1260671615600586,
+      "learning_rate": 4.570684641818663e-05,
+      "loss": 1.3839,
+      "step": 16251
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0426011085510254,
+      "learning_rate": 4.57065820777161e-05,
+      "loss": 1.3468,
+      "step": 16252
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.836783766746521,
+      "learning_rate": 4.570631773724557e-05,
+      "loss": 1.3549,
+      "step": 16253
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5297083854675293,
+      "learning_rate": 4.570605339677505e-05,
+      "loss": 1.3826,
+      "step": 16254
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6615781784057617,
+      "learning_rate": 4.570578905630452e-05,
+      "loss": 1.3511,
+      "step": 16255
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9284908771514893,
+      "learning_rate": 4.5705524715834e-05,
+      "loss": 1.4788,
+      "step": 16256
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3364663124084473,
+      "learning_rate": 4.5705260375363465e-05,
+      "loss": 1.2986,
+      "step": 16257
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.876721978187561,
+      "learning_rate": 4.5704996034892946e-05,
+      "loss": 1.3039,
+      "step": 16258
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.913455605506897,
+      "learning_rate": 4.570473169442242e-05,
+      "loss": 1.4981,
+      "step": 16259
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5901235342025757,
+      "learning_rate": 4.570446735395189e-05,
+      "loss": 1.3884,
+      "step": 16260
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8196395635604858,
+      "learning_rate": 4.5704203013481366e-05,
+      "loss": 1.4664,
+      "step": 16261
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7734668254852295,
+      "learning_rate": 4.570393867301084e-05,
+      "loss": 1.2119,
+      "step": 16262
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2161128520965576,
+      "learning_rate": 4.5703674332540313e-05,
+      "loss": 1.1718,
+      "step": 16263
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0142621994018555,
+      "learning_rate": 4.570340999206979e-05,
+      "loss": 1.4034,
+      "step": 16264
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3237709999084473,
+      "learning_rate": 4.570314565159926e-05,
+      "loss": 1.4893,
+      "step": 16265
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1584854125976562,
+      "learning_rate": 4.5702881311128734e-05,
+      "loss": 1.3946,
+      "step": 16266
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5881540775299072,
+      "learning_rate": 4.570261697065821e-05,
+      "loss": 1.1519,
+      "step": 16267
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.999786615371704,
+      "learning_rate": 4.570235263018768e-05,
+      "loss": 1.4824,
+      "step": 16268
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7721296548843384,
+      "learning_rate": 4.570208828971716e-05,
+      "loss": 1.3732,
+      "step": 16269
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8257356882095337,
+      "learning_rate": 4.570182394924663e-05,
+      "loss": 1.456,
+      "step": 16270
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6340601444244385,
+      "learning_rate": 4.570155960877611e-05,
+      "loss": 1.4485,
+      "step": 16271
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8558642864227295,
+      "learning_rate": 4.5701295268305576e-05,
+      "loss": 1.3,
+      "step": 16272
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.505875587463379,
+      "learning_rate": 4.5701030927835056e-05,
+      "loss": 1.3544,
+      "step": 16273
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7382811307907104,
+      "learning_rate": 4.570076658736453e-05,
+      "loss": 1.4426,
+      "step": 16274
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.360866069793701,
+      "learning_rate": 4.5700502246894e-05,
+      "loss": 1.5031,
+      "step": 16275
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.277642250061035,
+      "learning_rate": 4.570023790642348e-05,
+      "loss": 1.2534,
+      "step": 16276
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.709882378578186,
+      "learning_rate": 4.569997356595295e-05,
+      "loss": 1.5282,
+      "step": 16277
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7140581607818604,
+      "learning_rate": 4.5699709225482424e-05,
+      "loss": 1.4003,
+      "step": 16278
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.679403305053711,
+      "learning_rate": 4.56994448850119e-05,
+      "loss": 1.3932,
+      "step": 16279
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1720879077911377,
+      "learning_rate": 4.569918054454137e-05,
+      "loss": 1.374,
+      "step": 16280
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7441657781600952,
+      "learning_rate": 4.5698916204070845e-05,
+      "loss": 1.3571,
+      "step": 16281
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.784628987312317,
+      "learning_rate": 4.569865186360032e-05,
+      "loss": 1.4665,
+      "step": 16282
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1168742179870605,
+      "learning_rate": 4.569838752312979e-05,
+      "loss": 1.3606,
+      "step": 16283
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3041434288024902,
+      "learning_rate": 4.569812318265927e-05,
+      "loss": 1.6637,
+      "step": 16284
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4758795499801636,
+      "learning_rate": 4.569785884218874e-05,
+      "loss": 1.3173,
+      "step": 16285
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6599050760269165,
+      "learning_rate": 4.569759450171822e-05,
+      "loss": 1.3948,
+      "step": 16286
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3553245067596436,
+      "learning_rate": 4.5697330161247686e-05,
+      "loss": 1.2113,
+      "step": 16287
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8376092910766602,
+      "learning_rate": 4.569706582077716e-05,
+      "loss": 1.5137,
+      "step": 16288
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5649561882019043,
+      "learning_rate": 4.569680148030664e-05,
+      "loss": 1.2075,
+      "step": 16289
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8783177137374878,
+      "learning_rate": 4.569653713983611e-05,
+      "loss": 1.5196,
+      "step": 16290
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9127190113067627,
+      "learning_rate": 4.569627279936559e-05,
+      "loss": 1.4611,
+      "step": 16291
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.837924599647522,
+      "learning_rate": 4.5696008458895054e-05,
+      "loss": 1.4983,
+      "step": 16292
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.619871735572815,
+      "learning_rate": 4.5695744118424534e-05,
+      "loss": 1.4439,
+      "step": 16293
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6071758270263672,
+      "learning_rate": 4.569547977795401e-05,
+      "loss": 1.3083,
+      "step": 16294
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.177852153778076,
+      "learning_rate": 4.569521543748348e-05,
+      "loss": 1.4048,
+      "step": 16295
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8883540630340576,
+      "learning_rate": 4.5694951097012955e-05,
+      "loss": 1.4118,
+      "step": 16296
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9734550714492798,
+      "learning_rate": 4.569468675654243e-05,
+      "loss": 1.3661,
+      "step": 16297
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5457093715667725,
+      "learning_rate": 4.56944224160719e-05,
+      "loss": 1.2405,
+      "step": 16298
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.071504592895508,
+      "learning_rate": 4.5694158075601376e-05,
+      "loss": 1.4516,
+      "step": 16299
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7914119958877563,
+      "learning_rate": 4.569389373513085e-05,
+      "loss": 1.4958,
+      "step": 16300
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6464451551437378,
+      "learning_rate": 4.569362939466032e-05,
+      "loss": 1.421,
+      "step": 16301
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9509268999099731,
+      "learning_rate": 4.5693365054189796e-05,
+      "loss": 1.4023,
+      "step": 16302
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.123974084854126,
+      "learning_rate": 4.569310071371927e-05,
+      "loss": 1.4032,
+      "step": 16303
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6739609241485596,
+      "learning_rate": 4.569283637324875e-05,
+      "loss": 1.3285,
+      "step": 16304
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6959161758422852,
+      "learning_rate": 4.569257203277822e-05,
+      "loss": 1.375,
+      "step": 16305
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3925923109054565,
+      "learning_rate": 4.56923076923077e-05,
+      "loss": 2.4912,
+      "step": 16306
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.758972406387329,
+      "learning_rate": 4.5692043351837164e-05,
+      "loss": 1.3094,
+      "step": 16307
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7217273712158203,
+      "learning_rate": 4.5691779011366645e-05,
+      "loss": 1.243,
+      "step": 16308
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8733545541763306,
+      "learning_rate": 4.569151467089612e-05,
+      "loss": 1.4981,
+      "step": 16309
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.744560956954956,
+      "learning_rate": 4.569125033042559e-05,
+      "loss": 1.3833,
+      "step": 16310
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.764961838722229,
+      "learning_rate": 4.5690985989955065e-05,
+      "loss": 1.4974,
+      "step": 16311
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5029349327087402,
+      "learning_rate": 4.569072164948454e-05,
+      "loss": 1.3788,
+      "step": 16312
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.041353940963745,
+      "learning_rate": 4.569045730901401e-05,
+      "loss": 1.4336,
+      "step": 16313
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.764343023300171,
+      "learning_rate": 4.5690192968543486e-05,
+      "loss": 1.427,
+      "step": 16314
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2074737548828125,
+      "learning_rate": 4.568992862807296e-05,
+      "loss": 1.3946,
+      "step": 16315
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6437633037567139,
+      "learning_rate": 4.568966428760243e-05,
+      "loss": 1.3597,
+      "step": 16316
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.470082402229309,
+      "learning_rate": 4.568939994713191e-05,
+      "loss": 1.3276,
+      "step": 16317
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6984074115753174,
+      "learning_rate": 4.568913560666138e-05,
+      "loss": 1.2558,
+      "step": 16318
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6947884559631348,
+      "learning_rate": 4.568887126619086e-05,
+      "loss": 1.3923,
+      "step": 16319
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9688082933425903,
+      "learning_rate": 4.568860692572033e-05,
+      "loss": 1.3452,
+      "step": 16320
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4591121673583984,
+      "learning_rate": 4.568834258524981e-05,
+      "loss": 1.3724,
+      "step": 16321
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4130773544311523,
+      "learning_rate": 4.5688078244779275e-05,
+      "loss": 1.2404,
+      "step": 16322
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6914063692092896,
+      "learning_rate": 4.5687813904308755e-05,
+      "loss": 1.3675,
+      "step": 16323
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5077102184295654,
+      "learning_rate": 4.568754956383823e-05,
+      "loss": 1.3579,
+      "step": 16324
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.469119071960449,
+      "learning_rate": 4.5687285223367695e-05,
+      "loss": 1.3877,
+      "step": 16325
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5737544298171997,
+      "learning_rate": 4.5687020882897176e-05,
+      "loss": 1.9704,
+      "step": 16326
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7760714292526245,
+      "learning_rate": 4.568675654242664e-05,
+      "loss": 1.4225,
+      "step": 16327
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5334153175354004,
+      "learning_rate": 4.568649220195612e-05,
+      "loss": 1.5054,
+      "step": 16328
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6233112812042236,
+      "learning_rate": 4.5686227861485596e-05,
+      "loss": 1.4123,
+      "step": 16329
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3361258506774902,
+      "learning_rate": 4.568596352101507e-05,
+      "loss": 1.4708,
+      "step": 16330
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0036873817443848,
+      "learning_rate": 4.5685699180544543e-05,
+      "loss": 1.5249,
+      "step": 16331
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.681408405303955,
+      "learning_rate": 4.568543484007402e-05,
+      "loss": 1.4527,
+      "step": 16332
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6772795915603638,
+      "learning_rate": 4.568517049960349e-05,
+      "loss": 1.4499,
+      "step": 16333
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0531606674194336,
+      "learning_rate": 4.5684906159132964e-05,
+      "loss": 1.4974,
+      "step": 16334
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5821857452392578,
+      "learning_rate": 4.568464181866244e-05,
+      "loss": 1.2812,
+      "step": 16335
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8293519020080566,
+      "learning_rate": 4.568437747819191e-05,
+      "loss": 1.3081,
+      "step": 16336
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.654768943786621,
+      "learning_rate": 4.5684113137721385e-05,
+      "loss": 1.3484,
+      "step": 16337
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.406965494155884,
+      "learning_rate": 4.568384879725086e-05,
+      "loss": 1.4779,
+      "step": 16338
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9757425785064697,
+      "learning_rate": 4.568358445678034e-05,
+      "loss": 1.5855,
+      "step": 16339
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4772191047668457,
+      "learning_rate": 4.5683320116309806e-05,
+      "loss": 1.236,
+      "step": 16340
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.643491268157959,
+      "learning_rate": 4.5683055775839286e-05,
+      "loss": 1.5124,
+      "step": 16341
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9717378616333008,
+      "learning_rate": 4.568279143536875e-05,
+      "loss": 1.3623,
+      "step": 16342
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2650842666625977,
+      "learning_rate": 4.568252709489823e-05,
+      "loss": 1.4814,
+      "step": 16343
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7949939966201782,
+      "learning_rate": 4.568226275442771e-05,
+      "loss": 1.4973,
+      "step": 16344
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4462820291519165,
+      "learning_rate": 4.568199841395718e-05,
+      "loss": 2.255,
+      "step": 16345
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.207068681716919,
+      "learning_rate": 4.5681734073486654e-05,
+      "loss": 1.4933,
+      "step": 16346
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4368380308151245,
+      "learning_rate": 4.568146973301613e-05,
+      "loss": 1.2291,
+      "step": 16347
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.954439640045166,
+      "learning_rate": 4.56812053925456e-05,
+      "loss": 1.3637,
+      "step": 16348
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4716527462005615,
+      "learning_rate": 4.5680941052075074e-05,
+      "loss": 1.2698,
+      "step": 16349
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5795412063598633,
+      "learning_rate": 4.568067671160455e-05,
+      "loss": 1.4067,
+      "step": 16350
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9270602464675903,
+      "learning_rate": 4.568041237113402e-05,
+      "loss": 1.4391,
+      "step": 16351
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7647258043289185,
+      "learning_rate": 4.5680148030663495e-05,
+      "loss": 1.369,
+      "step": 16352
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1432929039001465,
+      "learning_rate": 4.567988369019297e-05,
+      "loss": 1.3252,
+      "step": 16353
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5494918823242188,
+      "learning_rate": 4.567961934972245e-05,
+      "loss": 1.7894,
+      "step": 16354
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6935451030731201,
+      "learning_rate": 4.5679355009251916e-05,
+      "loss": 1.1828,
+      "step": 16355
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6355341672897339,
+      "learning_rate": 4.5679090668781396e-05,
+      "loss": 1.4038,
+      "step": 16356
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.621409296989441,
+      "learning_rate": 4.567882632831086e-05,
+      "loss": 1.5373,
+      "step": 16357
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.283311128616333,
+      "learning_rate": 4.5678561987840343e-05,
+      "loss": 1.5336,
+      "step": 16358
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9307376146316528,
+      "learning_rate": 4.567829764736982e-05,
+      "loss": 1.4021,
+      "step": 16359
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2339961528778076,
+      "learning_rate": 4.567803330689929e-05,
+      "loss": 1.574,
+      "step": 16360
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8957948684692383,
+      "learning_rate": 4.5677768966428764e-05,
+      "loss": 1.3806,
+      "step": 16361
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3964192867279053,
+      "learning_rate": 4.567750462595823e-05,
+      "loss": 1.5708,
+      "step": 16362
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5201966762542725,
+      "learning_rate": 4.567724028548771e-05,
+      "loss": 1.8339,
+      "step": 16363
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4375600814819336,
+      "learning_rate": 4.5676975945017185e-05,
+      "loss": 1.3693,
+      "step": 16364
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.541334629058838,
+      "learning_rate": 4.567671160454666e-05,
+      "loss": 1.319,
+      "step": 16365
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9925646781921387,
+      "learning_rate": 4.567644726407613e-05,
+      "loss": 1.4786,
+      "step": 16366
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7765700817108154,
+      "learning_rate": 4.5676182923605606e-05,
+      "loss": 1.3172,
+      "step": 16367
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5716345310211182,
+      "learning_rate": 4.567591858313508e-05,
+      "loss": 1.2372,
+      "step": 16368
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4862531423568726,
+      "learning_rate": 4.567565424266456e-05,
+      "loss": 1.262,
+      "step": 16369
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6029987335205078,
+      "learning_rate": 4.5675389902194026e-05,
+      "loss": 1.7648,
+      "step": 16370
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4283058643341064,
+      "learning_rate": 4.56751255617235e-05,
+      "loss": 1.3859,
+      "step": 16371
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.931097149848938,
+      "learning_rate": 4.567486122125297e-05,
+      "loss": 1.4603,
+      "step": 16372
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1612038612365723,
+      "learning_rate": 4.567459688078245e-05,
+      "loss": 1.4961,
+      "step": 16373
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1540095806121826,
+      "learning_rate": 4.567433254031193e-05,
+      "loss": 1.6019,
+      "step": 16374
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.842850685119629,
+      "learning_rate": 4.5674068199841394e-05,
+      "loss": 1.4104,
+      "step": 16375
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.909289836883545,
+      "learning_rate": 4.5673803859370874e-05,
+      "loss": 1.3353,
+      "step": 16376
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.627746343612671,
+      "learning_rate": 4.567353951890034e-05,
+      "loss": 1.9211,
+      "step": 16377
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4167327880859375,
+      "learning_rate": 4.567327517842982e-05,
+      "loss": 1.3124,
+      "step": 16378
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8161542415618896,
+      "learning_rate": 4.5673010837959295e-05,
+      "loss": 1.4096,
+      "step": 16379
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9379384517669678,
+      "learning_rate": 4.567274649748877e-05,
+      "loss": 1.4489,
+      "step": 16380
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4868190288543701,
+      "learning_rate": 4.567248215701824e-05,
+      "loss": 1.2735,
+      "step": 16381
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.75896418094635,
+      "learning_rate": 4.5672217816547716e-05,
+      "loss": 1.1995,
+      "step": 16382
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.817190408706665,
+      "learning_rate": 4.567195347607719e-05,
+      "loss": 1.3065,
+      "step": 16383
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5966559648513794,
+      "learning_rate": 4.567168913560666e-05,
+      "loss": 1.4673,
+      "step": 16384
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6797072887420654,
+      "learning_rate": 4.5671424795136137e-05,
+      "loss": 1.3429,
+      "step": 16385
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8917661905288696,
+      "learning_rate": 4.567116045466561e-05,
+      "loss": 1.3395,
+      "step": 16386
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.697922468185425,
+      "learning_rate": 4.5670896114195084e-05,
+      "loss": 1.6351,
+      "step": 16387
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.292274236679077,
+      "learning_rate": 4.567063177372456e-05,
+      "loss": 1.439,
+      "step": 16388
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0351624488830566,
+      "learning_rate": 4.567036743325404e-05,
+      "loss": 1.4543,
+      "step": 16389
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4642555713653564,
+      "learning_rate": 4.5670103092783504e-05,
+      "loss": 1.2508,
+      "step": 16390
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5638556480407715,
+      "learning_rate": 4.5669838752312985e-05,
+      "loss": 1.3539,
+      "step": 16391
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5669320821762085,
+      "learning_rate": 4.566957441184245e-05,
+      "loss": 1.274,
+      "step": 16392
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5758565664291382,
+      "learning_rate": 4.566931007137193e-05,
+      "loss": 1.3826,
+      "step": 16393
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6988530158996582,
+      "learning_rate": 4.5669045730901405e-05,
+      "loss": 1.4267,
+      "step": 16394
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7230123281478882,
+      "learning_rate": 4.566878139043088e-05,
+      "loss": 1.3251,
+      "step": 16395
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4769141674041748,
+      "learning_rate": 4.566851704996035e-05,
+      "loss": 1.4413,
+      "step": 16396
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4033252000808716,
+      "learning_rate": 4.566825270948982e-05,
+      "loss": 1.3725,
+      "step": 16397
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7736477851867676,
+      "learning_rate": 4.56679883690193e-05,
+      "loss": 1.328,
+      "step": 16398
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.690707802772522,
+      "learning_rate": 4.566772402854877e-05,
+      "loss": 1.2109,
+      "step": 16399
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5154294967651367,
+      "learning_rate": 4.566745968807825e-05,
+      "loss": 1.1376,
+      "step": 16400
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6818519830703735,
+      "learning_rate": 4.566719534760772e-05,
+      "loss": 1.3757,
+      "step": 16401
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7241986989974976,
+      "learning_rate": 4.5666931007137194e-05,
+      "loss": 1.1536,
+      "step": 16402
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5580531358718872,
+      "learning_rate": 4.566666666666667e-05,
+      "loss": 1.4288,
+      "step": 16403
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5946635007858276,
+      "learning_rate": 4.566640232619615e-05,
+      "loss": 1.3544,
+      "step": 16404
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3597605228424072,
+      "learning_rate": 4.5666137985725615e-05,
+      "loss": 1.4035,
+      "step": 16405
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8840070962905884,
+      "learning_rate": 4.566587364525509e-05,
+      "loss": 1.352,
+      "step": 16406
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8316022157669067,
+      "learning_rate": 4.566560930478456e-05,
+      "loss": 1.3383,
+      "step": 16407
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1722943782806396,
+      "learning_rate": 4.5665344964314035e-05,
+      "loss": 1.4038,
+      "step": 16408
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9504727125167847,
+      "learning_rate": 4.5665080623843516e-05,
+      "loss": 1.2933,
+      "step": 16409
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9877907037734985,
+      "learning_rate": 4.566481628337298e-05,
+      "loss": 1.6231,
+      "step": 16410
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2144012451171875,
+      "learning_rate": 4.566455194290246e-05,
+      "loss": 1.3672,
+      "step": 16411
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8216363191604614,
+      "learning_rate": 4.566428760243193e-05,
+      "loss": 1.5345,
+      "step": 16412
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1841461658477783,
+      "learning_rate": 4.566402326196141e-05,
+      "loss": 1.4821,
+      "step": 16413
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.883423924446106,
+      "learning_rate": 4.5663758921490884e-05,
+      "loss": 1.3854,
+      "step": 16414
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6026932001113892,
+      "learning_rate": 4.566349458102036e-05,
+      "loss": 1.3565,
+      "step": 16415
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4889366626739502,
+      "learning_rate": 4.566323024054983e-05,
+      "loss": 1.37,
+      "step": 16416
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6975460052490234,
+      "learning_rate": 4.5662965900079304e-05,
+      "loss": 1.2561,
+      "step": 16417
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.561309576034546,
+      "learning_rate": 4.566270155960878e-05,
+      "loss": 1.2352,
+      "step": 16418
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8447463512420654,
+      "learning_rate": 4.566243721913825e-05,
+      "loss": 1.409,
+      "step": 16419
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6718225479125977,
+      "learning_rate": 4.5662172878667725e-05,
+      "loss": 1.3612,
+      "step": 16420
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6202144622802734,
+      "learning_rate": 4.56619085381972e-05,
+      "loss": 1.3396,
+      "step": 16421
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6963458061218262,
+      "learning_rate": 4.566164419772667e-05,
+      "loss": 1.47,
+      "step": 16422
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2671148777008057,
+      "learning_rate": 4.5661379857256146e-05,
+      "loss": 1.5476,
+      "step": 16423
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.680410623550415,
+      "learning_rate": 4.5661115516785626e-05,
+      "loss": 1.3972,
+      "step": 16424
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.049281597137451,
+      "learning_rate": 4.566085117631509e-05,
+      "loss": 1.4899,
+      "step": 16425
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6885573863983154,
+      "learning_rate": 4.566058683584457e-05,
+      "loss": 1.4128,
+      "step": 16426
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8243942260742188,
+      "learning_rate": 4.566032249537404e-05,
+      "loss": 1.3136,
+      "step": 16427
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9557842016220093,
+      "learning_rate": 4.566005815490352e-05,
+      "loss": 1.4713,
+      "step": 16428
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7601431608200073,
+      "learning_rate": 4.5659793814432994e-05,
+      "loss": 1.2818,
+      "step": 16429
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8972558975219727,
+      "learning_rate": 4.565952947396247e-05,
+      "loss": 1.3301,
+      "step": 16430
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.569258689880371,
+      "learning_rate": 4.565926513349194e-05,
+      "loss": 1.1652,
+      "step": 16431
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.135882616043091,
+      "learning_rate": 4.5659000793021415e-05,
+      "loss": 1.4936,
+      "step": 16432
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8132150173187256,
+      "learning_rate": 4.565873645255089e-05,
+      "loss": 1.4548,
+      "step": 16433
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6133300065994263,
+      "learning_rate": 4.565847211208036e-05,
+      "loss": 1.3931,
+      "step": 16434
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0772812366485596,
+      "learning_rate": 4.5658207771609835e-05,
+      "loss": 1.4732,
+      "step": 16435
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8837382793426514,
+      "learning_rate": 4.565794343113931e-05,
+      "loss": 1.443,
+      "step": 16436
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2349774837493896,
+      "learning_rate": 4.565767909066878e-05,
+      "loss": 1.4962,
+      "step": 16437
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9385946989059448,
+      "learning_rate": 4.5657414750198256e-05,
+      "loss": 1.5461,
+      "step": 16438
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.116213321685791,
+      "learning_rate": 4.5657150409727737e-05,
+      "loss": 1.3747,
+      "step": 16439
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2313320636749268,
+      "learning_rate": 4.56568860692572e-05,
+      "loss": 1.4615,
+      "step": 16440
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8453924655914307,
+      "learning_rate": 4.5656621728786684e-05,
+      "loss": 1.4307,
+      "step": 16441
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.200076103210449,
+      "learning_rate": 4.565635738831615e-05,
+      "loss": 1.5235,
+      "step": 16442
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.061103343963623,
+      "learning_rate": 4.5656093047845624e-05,
+      "loss": 1.5278,
+      "step": 16443
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9227162599563599,
+      "learning_rate": 4.5655828707375104e-05,
+      "loss": 1.3126,
+      "step": 16444
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5239436626434326,
+      "learning_rate": 4.565556436690457e-05,
+      "loss": 1.4564,
+      "step": 16445
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8781023025512695,
+      "learning_rate": 4.565530002643405e-05,
+      "loss": 1.4463,
+      "step": 16446
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7281343936920166,
+      "learning_rate": 4.565503568596352e-05,
+      "loss": 1.4106,
+      "step": 16447
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4616751670837402,
+      "learning_rate": 4.5654771345493e-05,
+      "loss": 1.3112,
+      "step": 16448
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5772972106933594,
+      "learning_rate": 4.565450700502247e-05,
+      "loss": 1.425,
+      "step": 16449
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2268025875091553,
+      "learning_rate": 4.5654242664551946e-05,
+      "loss": 1.3153,
+      "step": 16450
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.174039125442505,
+      "learning_rate": 4.565397832408142e-05,
+      "loss": 1.2247,
+      "step": 16451
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7163832187652588,
+      "learning_rate": 4.565371398361089e-05,
+      "loss": 1.3817,
+      "step": 16452
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.522145390510559,
+      "learning_rate": 4.5653449643140367e-05,
+      "loss": 1.284,
+      "step": 16453
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8228936195373535,
+      "learning_rate": 4.565318530266984e-05,
+      "loss": 1.3311,
+      "step": 16454
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.845857858657837,
+      "learning_rate": 4.5652920962199314e-05,
+      "loss": 1.4402,
+      "step": 16455
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8793307542800903,
+      "learning_rate": 4.565265662172879e-05,
+      "loss": 1.3698,
+      "step": 16456
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8590868711471558,
+      "learning_rate": 4.565239228125826e-05,
+      "loss": 1.5569,
+      "step": 16457
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6144921779632568,
+      "learning_rate": 4.5652127940787734e-05,
+      "loss": 1.5229,
+      "step": 16458
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2317960262298584,
+      "learning_rate": 4.5651863600317215e-05,
+      "loss": 1.2204,
+      "step": 16459
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4122518301010132,
+      "learning_rate": 4.565159925984668e-05,
+      "loss": 1.4382,
+      "step": 16460
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.727462887763977,
+      "learning_rate": 4.565133491937616e-05,
+      "loss": 1.3484,
+      "step": 16461
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.400019645690918,
+      "learning_rate": 4.565107057890563e-05,
+      "loss": 1.4398,
+      "step": 16462
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7884230613708496,
+      "learning_rate": 4.565080623843511e-05,
+      "loss": 1.4681,
+      "step": 16463
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8789576292037964,
+      "learning_rate": 4.565054189796458e-05,
+      "loss": 1.4381,
+      "step": 16464
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6755399703979492,
+      "learning_rate": 4.5650277557494056e-05,
+      "loss": 1.352,
+      "step": 16465
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9908239841461182,
+      "learning_rate": 4.565001321702353e-05,
+      "loss": 1.4844,
+      "step": 16466
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2115821838378906,
+      "learning_rate": 4.5649748876553e-05,
+      "loss": 1.4554,
+      "step": 16467
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.339563846588135,
+      "learning_rate": 4.564948453608248e-05,
+      "loss": 1.6133,
+      "step": 16468
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.43784761428833,
+      "learning_rate": 4.564922019561195e-05,
+      "loss": 1.238,
+      "step": 16469
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4423757791519165,
+      "learning_rate": 4.5648955855141424e-05,
+      "loss": 1.3341,
+      "step": 16470
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.574953556060791,
+      "learning_rate": 4.56486915146709e-05,
+      "loss": 1.3021,
+      "step": 16471
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8630502223968506,
+      "learning_rate": 4.564842717420037e-05,
+      "loss": 1.4272,
+      "step": 16472
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0259218215942383,
+      "learning_rate": 4.5648162833729845e-05,
+      "loss": 1.538,
+      "step": 16473
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1543354988098145,
+      "learning_rate": 4.5647898493259325e-05,
+      "loss": 1.4858,
+      "step": 16474
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6379824876785278,
+      "learning_rate": 4.564763415278879e-05,
+      "loss": 1.4337,
+      "step": 16475
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.427196741104126,
+      "learning_rate": 4.564736981231827e-05,
+      "loss": 1.2552,
+      "step": 16476
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5818055868148804,
+      "learning_rate": 4.564710547184774e-05,
+      "loss": 1.4501,
+      "step": 16477
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5472774505615234,
+      "learning_rate": 4.564684113137722e-05,
+      "loss": 1.3487,
+      "step": 16478
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.0230937004089355,
+      "learning_rate": 4.564657679090669e-05,
+      "loss": 1.5815,
+      "step": 16479
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9980250597000122,
+      "learning_rate": 4.564631245043616e-05,
+      "loss": 1.3783,
+      "step": 16480
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6464861631393433,
+      "learning_rate": 4.564604810996564e-05,
+      "loss": 1.245,
+      "step": 16481
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8112484216690063,
+      "learning_rate": 4.564578376949511e-05,
+      "loss": 1.3621,
+      "step": 16482
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8009631633758545,
+      "learning_rate": 4.564551942902459e-05,
+      "loss": 1.5766,
+      "step": 16483
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8820915222167969,
+      "learning_rate": 4.564525508855406e-05,
+      "loss": 1.4507,
+      "step": 16484
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.236478328704834,
+      "learning_rate": 4.5644990748083534e-05,
+      "loss": 1.4223,
+      "step": 16485
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9645155668258667,
+      "learning_rate": 4.564472640761301e-05,
+      "loss": 2.0367,
+      "step": 16486
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9374306201934814,
+      "learning_rate": 4.564446206714248e-05,
+      "loss": 1.3079,
+      "step": 16487
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9225722551345825,
+      "learning_rate": 4.5644197726671955e-05,
+      "loss": 1.4225,
+      "step": 16488
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4641190767288208,
+      "learning_rate": 4.564393338620143e-05,
+      "loss": 1.2544,
+      "step": 16489
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2743756771087646,
+      "learning_rate": 4.56436690457309e-05,
+      "loss": 1.5992,
+      "step": 16490
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.727244257926941,
+      "learning_rate": 4.5643404705260376e-05,
+      "loss": 1.2682,
+      "step": 16491
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0733842849731445,
+      "learning_rate": 4.564314036478985e-05,
+      "loss": 1.5738,
+      "step": 16492
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9333744049072266,
+      "learning_rate": 4.564287602431932e-05,
+      "loss": 1.3966,
+      "step": 16493
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8475334644317627,
+      "learning_rate": 4.56426116838488e-05,
+      "loss": 1.4541,
+      "step": 16494
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5113716125488281,
+      "learning_rate": 4.564234734337827e-05,
+      "loss": 1.3706,
+      "step": 16495
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0344669818878174,
+      "learning_rate": 4.564208300290775e-05,
+      "loss": 1.4317,
+      "step": 16496
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4899346828460693,
+      "learning_rate": 4.564181866243722e-05,
+      "loss": 1.3079,
+      "step": 16497
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7908565998077393,
+      "learning_rate": 4.56415543219667e-05,
+      "loss": 1.4852,
+      "step": 16498
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.026344060897827,
+      "learning_rate": 4.564128998149617e-05,
+      "loss": 1.3467,
+      "step": 16499
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3067307472229004,
+      "learning_rate": 4.5641025641025645e-05,
+      "loss": 1.3852,
+      "step": 16500
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8378138542175293,
+      "learning_rate": 4.564076130055512e-05,
+      "loss": 1.3094,
+      "step": 16501
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9654988050460815,
+      "learning_rate": 4.564049696008459e-05,
+      "loss": 1.3089,
+      "step": 16502
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8915174007415771,
+      "learning_rate": 4.5640232619614065e-05,
+      "loss": 1.4544,
+      "step": 16503
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6756322383880615,
+      "learning_rate": 4.563996827914354e-05,
+      "loss": 1.31,
+      "step": 16504
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6556313037872314,
+      "learning_rate": 4.563970393867301e-05,
+      "loss": 1.4392,
+      "step": 16505
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8495486974716187,
+      "learning_rate": 4.5639439598202486e-05,
+      "loss": 1.5525,
+      "step": 16506
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8250973224639893,
+      "learning_rate": 4.563917525773196e-05,
+      "loss": 1.5908,
+      "step": 16507
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.783773422241211,
+      "learning_rate": 4.563891091726143e-05,
+      "loss": 1.3005,
+      "step": 16508
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6697906255722046,
+      "learning_rate": 4.5638646576790914e-05,
+      "loss": 1.4243,
+      "step": 16509
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.896185040473938,
+      "learning_rate": 4.563838223632038e-05,
+      "loss": 1.5107,
+      "step": 16510
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6772081851959229,
+      "learning_rate": 4.563811789584986e-05,
+      "loss": 1.3483,
+      "step": 16511
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5620445013046265,
+      "learning_rate": 4.563785355537933e-05,
+      "loss": 1.088,
+      "step": 16512
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4066507816314697,
+      "learning_rate": 4.563758921490881e-05,
+      "loss": 1.4891,
+      "step": 16513
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.544113278388977,
+      "learning_rate": 4.563732487443828e-05,
+      "loss": 1.2507,
+      "step": 16514
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.809393286705017,
+      "learning_rate": 4.563706053396775e-05,
+      "loss": 1.4925,
+      "step": 16515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4133065938949585,
+      "learning_rate": 4.563679619349723e-05,
+      "loss": 1.231,
+      "step": 16516
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.81486976146698,
+      "learning_rate": 4.5636531853026695e-05,
+      "loss": 1.4696,
+      "step": 16517
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8991923332214355,
+      "learning_rate": 4.5636267512556176e-05,
+      "loss": 1.5024,
+      "step": 16518
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.328268527984619,
+      "learning_rate": 4.563600317208565e-05,
+      "loss": 1.6788,
+      "step": 16519
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1240224838256836,
+      "learning_rate": 4.563573883161512e-05,
+      "loss": 1.3724,
+      "step": 16520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8366026878356934,
+      "learning_rate": 4.5635474491144596e-05,
+      "loss": 1.2944,
+      "step": 16521
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8576103448867798,
+      "learning_rate": 4.563521015067407e-05,
+      "loss": 1.4295,
+      "step": 16522
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.879009485244751,
+      "learning_rate": 4.5634945810203544e-05,
+      "loss": 1.4912,
+      "step": 16523
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.481743812561035,
+      "learning_rate": 4.563468146973302e-05,
+      "loss": 1.5749,
+      "step": 16524
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.64753794670105,
+      "learning_rate": 4.563441712926249e-05,
+      "loss": 1.4546,
+      "step": 16525
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8227264881134033,
+      "learning_rate": 4.5634152788791964e-05,
+      "loss": 1.362,
+      "step": 16526
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6819802522659302,
+      "learning_rate": 4.563388844832144e-05,
+      "loss": 1.3798,
+      "step": 16527
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9964489936828613,
+      "learning_rate": 4.563362410785091e-05,
+      "loss": 1.482,
+      "step": 16528
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.783238172531128,
+      "learning_rate": 4.563335976738039e-05,
+      "loss": 1.5362,
+      "step": 16529
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.105823278427124,
+      "learning_rate": 4.563309542690986e-05,
+      "loss": 1.3269,
+      "step": 16530
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.048978805541992,
+      "learning_rate": 4.563283108643934e-05,
+      "loss": 1.4728,
+      "step": 16531
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.044947624206543,
+      "learning_rate": 4.5632566745968806e-05,
+      "loss": 1.3431,
+      "step": 16532
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.47638201713562,
+      "learning_rate": 4.5632302405498286e-05,
+      "loss": 1.5619,
+      "step": 16533
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9379974603652954,
+      "learning_rate": 4.563203806502776e-05,
+      "loss": 1.4829,
+      "step": 16534
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6171519756317139,
+      "learning_rate": 4.563177372455723e-05,
+      "loss": 1.3664,
+      "step": 16535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.387555718421936,
+      "learning_rate": 4.563150938408671e-05,
+      "loss": 1.2262,
+      "step": 16536
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6593877077102661,
+      "learning_rate": 4.563124504361618e-05,
+      "loss": 1.4019,
+      "step": 16537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8005497455596924,
+      "learning_rate": 4.5630980703145654e-05,
+      "loss": 1.6334,
+      "step": 16538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9349658489227295,
+      "learning_rate": 4.563071636267513e-05,
+      "loss": 1.404,
+      "step": 16539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6817697286605835,
+      "learning_rate": 4.56304520222046e-05,
+      "loss": 1.4322,
+      "step": 16540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8776048421859741,
+      "learning_rate": 4.5630187681734075e-05,
+      "loss": 1.4927,
+      "step": 16541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.178567409515381,
+      "learning_rate": 4.562992334126355e-05,
+      "loss": 1.506,
+      "step": 16542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9612107276916504,
+      "learning_rate": 4.562965900079302e-05,
+      "loss": 1.4488,
+      "step": 16543
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.596620559692383,
+      "learning_rate": 4.56293946603225e-05,
+      "loss": 1.5153,
+      "step": 16544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4943623542785645,
+      "learning_rate": 4.562913031985197e-05,
+      "loss": 1.2932,
+      "step": 16545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1418118476867676,
+      "learning_rate": 4.562886597938145e-05,
+      "loss": 1.3734,
+      "step": 16546
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.323033094406128,
+      "learning_rate": 4.5628601638910916e-05,
+      "loss": 1.6708,
+      "step": 16547
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.644586443901062,
+      "learning_rate": 4.5628337298440396e-05,
+      "loss": 1.3218,
+      "step": 16548
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4165306091308594,
+      "learning_rate": 4.562807295796987e-05,
+      "loss": 1.2163,
+      "step": 16549
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8368877172470093,
+      "learning_rate": 4.5627808617499344e-05,
+      "loss": 1.4121,
+      "step": 16550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4896997213363647,
+      "learning_rate": 4.562754427702882e-05,
+      "loss": 1.3099,
+      "step": 16551
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.811509132385254,
+      "learning_rate": 4.5627279936558284e-05,
+      "loss": 1.5364,
+      "step": 16552
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.919276475906372,
+      "learning_rate": 4.5627015596087764e-05,
+      "loss": 1.5089,
+      "step": 16553
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2790279388427734,
+      "learning_rate": 4.562675125561724e-05,
+      "loss": 1.6038,
+      "step": 16554
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1118783950805664,
+      "learning_rate": 4.562648691514671e-05,
+      "loss": 1.438,
+      "step": 16555
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.386079788208008,
+      "learning_rate": 4.5626222574676185e-05,
+      "loss": 1.6022,
+      "step": 16556
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.245514154434204,
+      "learning_rate": 4.562595823420566e-05,
+      "loss": 1.269,
+      "step": 16557
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6030240058898926,
+      "learning_rate": 4.562569389373513e-05,
+      "loss": 1.342,
+      "step": 16558
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5529155731201172,
+      "learning_rate": 4.562542955326461e-05,
+      "loss": 1.3261,
+      "step": 16559
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2811431884765625,
+      "learning_rate": 4.562516521279408e-05,
+      "loss": 1.4486,
+      "step": 16560
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6855239868164062,
+      "learning_rate": 4.562490087232355e-05,
+      "loss": 1.3942,
+      "step": 16561
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8818950653076172,
+      "learning_rate": 4.5624636531853026e-05,
+      "loss": 1.4463,
+      "step": 16562
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.50685715675354,
+      "learning_rate": 4.56243721913825e-05,
+      "loss": 1.1564,
+      "step": 16563
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5705188512802124,
+      "learning_rate": 4.562410785091198e-05,
+      "loss": 1.2655,
+      "step": 16564
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.287071704864502,
+      "learning_rate": 4.562384351044145e-05,
+      "loss": 1.4613,
+      "step": 16565
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6315908432006836,
+      "learning_rate": 4.562357916997093e-05,
+      "loss": 1.3339,
+      "step": 16566
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0825467109680176,
+      "learning_rate": 4.5623314829500394e-05,
+      "loss": 1.5473,
+      "step": 16567
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7344164848327637,
+      "learning_rate": 4.5623050489029875e-05,
+      "loss": 1.6987,
+      "step": 16568
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6228123903274536,
+      "learning_rate": 4.562278614855935e-05,
+      "loss": 1.3531,
+      "step": 16569
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.897552490234375,
+      "learning_rate": 4.562252180808882e-05,
+      "loss": 1.3662,
+      "step": 16570
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.372618317604065,
+      "learning_rate": 4.5622257467618295e-05,
+      "loss": 1.2102,
+      "step": 16571
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8532353639602661,
+      "learning_rate": 4.562199312714777e-05,
+      "loss": 1.4582,
+      "step": 16572
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5644903182983398,
+      "learning_rate": 4.562172878667724e-05,
+      "loss": 1.2979,
+      "step": 16573
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5469748973846436,
+      "learning_rate": 4.5621464446206716e-05,
+      "loss": 1.1503,
+      "step": 16574
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6627073287963867,
+      "learning_rate": 4.562120010573619e-05,
+      "loss": 1.2652,
+      "step": 16575
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9126936197280884,
+      "learning_rate": 4.562093576526566e-05,
+      "loss": 1.4308,
+      "step": 16576
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.741265892982483,
+      "learning_rate": 4.562067142479514e-05,
+      "loss": 1.3996,
+      "step": 16577
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.587022066116333,
+      "learning_rate": 4.562040708432461e-05,
+      "loss": 1.4633,
+      "step": 16578
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.313270330429077,
+      "learning_rate": 4.562014274385409e-05,
+      "loss": 1.5784,
+      "step": 16579
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.396541118621826,
+      "learning_rate": 4.561987840338356e-05,
+      "loss": 1.4614,
+      "step": 16580
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7430728673934937,
+      "learning_rate": 4.561961406291304e-05,
+      "loss": 1.4904,
+      "step": 16581
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.392386794090271,
+      "learning_rate": 4.5619349722442505e-05,
+      "loss": 1.2849,
+      "step": 16582
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7225881814956665,
+      "learning_rate": 4.5619085381971985e-05,
+      "loss": 1.4503,
+      "step": 16583
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6576192378997803,
+      "learning_rate": 4.561882104150146e-05,
+      "loss": 1.4581,
+      "step": 16584
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9444854259490967,
+      "learning_rate": 4.561855670103093e-05,
+      "loss": 1.3962,
+      "step": 16585
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.571518063545227,
+      "learning_rate": 4.5618292360560406e-05,
+      "loss": 1.2716,
+      "step": 16586
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.507123589515686,
+      "learning_rate": 4.561802802008987e-05,
+      "loss": 1.7151,
+      "step": 16587
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.554511070251465,
+      "learning_rate": 4.561776367961935e-05,
+      "loss": 1.3117,
+      "step": 16588
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9704457521438599,
+      "learning_rate": 4.5617499339148826e-05,
+      "loss": 1.1873,
+      "step": 16589
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4915047883987427,
+      "learning_rate": 4.56172349986783e-05,
+      "loss": 1.3171,
+      "step": 16590
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8236044645309448,
+      "learning_rate": 4.5616970658207774e-05,
+      "loss": 1.4879,
+      "step": 16591
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0884664058685303,
+      "learning_rate": 4.561670631773725e-05,
+      "loss": 1.4675,
+      "step": 16592
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7364133596420288,
+      "learning_rate": 4.561644197726672e-05,
+      "loss": 1.4776,
+      "step": 16593
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.521188497543335,
+      "learning_rate": 4.56161776367962e-05,
+      "loss": 1.2699,
+      "step": 16594
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5919289588928223,
+      "learning_rate": 4.561591329632567e-05,
+      "loss": 1.4389,
+      "step": 16595
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.095831871032715,
+      "learning_rate": 4.561564895585515e-05,
+      "loss": 1.3138,
+      "step": 16596
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6339644193649292,
+      "learning_rate": 4.5615384615384615e-05,
+      "loss": 1.1523,
+      "step": 16597
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6018370389938354,
+      "learning_rate": 4.561512027491409e-05,
+      "loss": 1.3076,
+      "step": 16598
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6461331844329834,
+      "learning_rate": 4.561485593444357e-05,
+      "loss": 1.3149,
+      "step": 16599
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.918498158454895,
+      "learning_rate": 4.5614591593973036e-05,
+      "loss": 1.3607,
+      "step": 16600
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5071029663085938,
+      "learning_rate": 4.5614327253502516e-05,
+      "loss": 1.3813,
+      "step": 16601
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0099923610687256,
+      "learning_rate": 4.561406291303198e-05,
+      "loss": 1.4934,
+      "step": 16602
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5076167583465576,
+      "learning_rate": 4.561379857256146e-05,
+      "loss": 1.6122,
+      "step": 16603
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6958954334259033,
+      "learning_rate": 4.561353423209094e-05,
+      "loss": 1.426,
+      "step": 16604
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7146025896072388,
+      "learning_rate": 4.561326989162041e-05,
+      "loss": 1.4835,
+      "step": 16605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.109851837158203,
+      "learning_rate": 4.5613005551149884e-05,
+      "loss": 1.5168,
+      "step": 16606
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8484700918197632,
+      "learning_rate": 4.561274121067936e-05,
+      "loss": 1.5079,
+      "step": 16607
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3845722675323486,
+      "learning_rate": 4.561247687020883e-05,
+      "loss": 1.5524,
+      "step": 16608
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5301947593688965,
+      "learning_rate": 4.5612212529738305e-05,
+      "loss": 1.2822,
+      "step": 16609
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3874151706695557,
+      "learning_rate": 4.561194818926778e-05,
+      "loss": 1.6499,
+      "step": 16610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.080217123031616,
+      "learning_rate": 4.561168384879725e-05,
+      "loss": 1.3125,
+      "step": 16611
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6075307130813599,
+      "learning_rate": 4.5611419508326725e-05,
+      "loss": 1.4281,
+      "step": 16612
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1375656127929688,
+      "learning_rate": 4.56111551678562e-05,
+      "loss": 1.4391,
+      "step": 16613
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1797738075256348,
+      "learning_rate": 4.561089082738568e-05,
+      "loss": 1.5283,
+      "step": 16614
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5975526571273804,
+      "learning_rate": 4.5610626486915146e-05,
+      "loss": 1.2801,
+      "step": 16615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5160843133926392,
+      "learning_rate": 4.5610362146444626e-05,
+      "loss": 1.2996,
+      "step": 16616
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.843131422996521,
+      "learning_rate": 4.561009780597409e-05,
+      "loss": 1.4443,
+      "step": 16617
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.401813268661499,
+      "learning_rate": 4.5609833465503573e-05,
+      "loss": 1.6583,
+      "step": 16618
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3317738771438599,
+      "learning_rate": 4.560956912503305e-05,
+      "loss": 1.1661,
+      "step": 16619
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5440597534179688,
+      "learning_rate": 4.560930478456252e-05,
+      "loss": 1.5745,
+      "step": 16620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.607619047164917,
+      "learning_rate": 4.5609040444091994e-05,
+      "loss": 1.4033,
+      "step": 16621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6902707815170288,
+      "learning_rate": 4.560877610362147e-05,
+      "loss": 1.3517,
+      "step": 16622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9772311449050903,
+      "learning_rate": 4.560851176315094e-05,
+      "loss": 1.4523,
+      "step": 16623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.019954204559326,
+      "learning_rate": 4.5608247422680415e-05,
+      "loss": 1.5157,
+      "step": 16624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.589826226234436,
+      "learning_rate": 4.560798308220989e-05,
+      "loss": 1.3859,
+      "step": 16625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.768821358680725,
+      "learning_rate": 4.560771874173936e-05,
+      "loss": 2.5439,
+      "step": 16626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.282135009765625,
+      "learning_rate": 4.5607454401268836e-05,
+      "loss": 1.1485,
+      "step": 16627
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6172471046447754,
+      "learning_rate": 4.560719006079831e-05,
+      "loss": 1.5601,
+      "step": 16628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6444977521896362,
+      "learning_rate": 4.560692572032779e-05,
+      "loss": 1.3644,
+      "step": 16629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3956172466278076,
+      "learning_rate": 4.5606661379857256e-05,
+      "loss": 1.4558,
+      "step": 16630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8818169832229614,
+      "learning_rate": 4.560639703938674e-05,
+      "loss": 1.4711,
+      "step": 16631
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6872199773788452,
+      "learning_rate": 4.5606132698916203e-05,
+      "loss": 1.3869,
+      "step": 16632
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5573647022247314,
+      "learning_rate": 4.560586835844568e-05,
+      "loss": 1.4092,
+      "step": 16633
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5613129138946533,
+      "learning_rate": 4.560560401797516e-05,
+      "loss": 1.2907,
+      "step": 16634
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8720625638961792,
+      "learning_rate": 4.5605339677504624e-05,
+      "loss": 1.3868,
+      "step": 16635
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6258044242858887,
+      "learning_rate": 4.5605075337034105e-05,
+      "loss": 1.2856,
+      "step": 16636
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.177186965942383,
+      "learning_rate": 4.560481099656357e-05,
+      "loss": 1.4813,
+      "step": 16637
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9106242656707764,
+      "learning_rate": 4.560454665609305e-05,
+      "loss": 1.4078,
+      "step": 16638
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5428452491760254,
+      "learning_rate": 4.5604282315622525e-05,
+      "loss": 1.4873,
+      "step": 16639
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.910261631011963,
+      "learning_rate": 4.5604017975152e-05,
+      "loss": 1.3733,
+      "step": 16640
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6770381927490234,
+      "learning_rate": 4.560375363468147e-05,
+      "loss": 1.4179,
+      "step": 16641
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.544281244277954,
+      "learning_rate": 4.5603489294210946e-05,
+      "loss": 1.224,
+      "step": 16642
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.818077802658081,
+      "learning_rate": 4.560322495374042e-05,
+      "loss": 1.458,
+      "step": 16643
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5123422145843506,
+      "learning_rate": 4.560296061326989e-05,
+      "loss": 1.2223,
+      "step": 16644
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8864657878875732,
+      "learning_rate": 4.560269627279937e-05,
+      "loss": 1.5283,
+      "step": 16645
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8995164632797241,
+      "learning_rate": 4.560243193232884e-05,
+      "loss": 1.4027,
+      "step": 16646
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7925297021865845,
+      "learning_rate": 4.5602167591858314e-05,
+      "loss": 1.3642,
+      "step": 16647
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4475303888320923,
+      "learning_rate": 4.560190325138779e-05,
+      "loss": 1.4912,
+      "step": 16648
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6305434703826904,
+      "learning_rate": 4.560163891091727e-05,
+      "loss": 1.4665,
+      "step": 16649
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2458062171936035,
+      "learning_rate": 4.5601374570446735e-05,
+      "loss": 1.2368,
+      "step": 16650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.716808557510376,
+      "learning_rate": 4.5601110229976215e-05,
+      "loss": 1.3562,
+      "step": 16651
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.972882866859436,
+      "learning_rate": 4.560084588950568e-05,
+      "loss": 1.4122,
+      "step": 16652
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7600336074829102,
+      "learning_rate": 4.560058154903516e-05,
+      "loss": 1.4303,
+      "step": 16653
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.232898235321045,
+      "learning_rate": 4.5600317208564636e-05,
+      "loss": 1.4554,
+      "step": 16654
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.010798692703247,
+      "learning_rate": 4.560005286809411e-05,
+      "loss": 1.3859,
+      "step": 16655
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6102365255355835,
+      "learning_rate": 4.559978852762358e-05,
+      "loss": 1.4278,
+      "step": 16656
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3402904272079468,
+      "learning_rate": 4.5599524187153056e-05,
+      "loss": 1.2455,
+      "step": 16657
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6462510824203491,
+      "learning_rate": 4.559925984668253e-05,
+      "loss": 1.1567,
+      "step": 16658
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.75315260887146,
+      "learning_rate": 4.5598995506212003e-05,
+      "loss": 1.4629,
+      "step": 16659
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9793940782546997,
+      "learning_rate": 4.559873116574148e-05,
+      "loss": 1.5517,
+      "step": 16660
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6364175081253052,
+      "learning_rate": 4.559846682527095e-05,
+      "loss": 1.3182,
+      "step": 16661
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7353148460388184,
+      "learning_rate": 4.5598202484800424e-05,
+      "loss": 1.3365,
+      "step": 16662
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.038299322128296,
+      "learning_rate": 4.55979381443299e-05,
+      "loss": 1.3911,
+      "step": 16663
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.824600100517273,
+      "learning_rate": 4.559767380385938e-05,
+      "loss": 1.3932,
+      "step": 16664
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8773192167282104,
+      "learning_rate": 4.5597409463388845e-05,
+      "loss": 1.4265,
+      "step": 16665
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.787424087524414,
+      "learning_rate": 4.5597145122918325e-05,
+      "loss": 1.3226,
+      "step": 16666
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4749839305877686,
+      "learning_rate": 4.559688078244779e-05,
+      "loss": 1.4026,
+      "step": 16667
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.702883243560791,
+      "learning_rate": 4.559661644197727e-05,
+      "loss": 1.3542,
+      "step": 16668
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.010655641555786,
+      "learning_rate": 4.5596352101506746e-05,
+      "loss": 1.4203,
+      "step": 16669
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0203380584716797,
+      "learning_rate": 4.559608776103621e-05,
+      "loss": 1.2684,
+      "step": 16670
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6243863105773926,
+      "learning_rate": 4.559582342056569e-05,
+      "loss": 1.2425,
+      "step": 16671
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7598742246627808,
+      "learning_rate": 4.559555908009516e-05,
+      "loss": 1.3809,
+      "step": 16672
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7925832271575928,
+      "learning_rate": 4.559529473962464e-05,
+      "loss": 1.3854,
+      "step": 16673
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6327695846557617,
+      "learning_rate": 4.5595030399154114e-05,
+      "loss": 1.3398,
+      "step": 16674
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7376660108566284,
+      "learning_rate": 4.559476605868359e-05,
+      "loss": 1.2945,
+      "step": 16675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.081646203994751,
+      "learning_rate": 4.559450171821306e-05,
+      "loss": 1.5494,
+      "step": 16676
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5015745162963867,
+      "learning_rate": 4.5594237377742534e-05,
+      "loss": 1.454,
+      "step": 16677
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.921079397201538,
+      "learning_rate": 4.559397303727201e-05,
+      "loss": 1.4017,
+      "step": 16678
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9940249919891357,
+      "learning_rate": 4.559370869680148e-05,
+      "loss": 1.4086,
+      "step": 16679
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6481496095657349,
+      "learning_rate": 4.5593444356330955e-05,
+      "loss": 2.557,
+      "step": 16680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.046736717224121,
+      "learning_rate": 4.559318001586043e-05,
+      "loss": 1.5031,
+      "step": 16681
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7336773872375488,
+      "learning_rate": 4.55929156753899e-05,
+      "loss": 1.4224,
+      "step": 16682
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1257150173187256,
+      "learning_rate": 4.5592651334919376e-05,
+      "loss": 1.4462,
+      "step": 16683
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7987184524536133,
+      "learning_rate": 4.5592386994448856e-05,
+      "loss": 1.3092,
+      "step": 16684
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4555110931396484,
+      "learning_rate": 4.559212265397832e-05,
+      "loss": 1.5052,
+      "step": 16685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1174542903900146,
+      "learning_rate": 4.5591858313507803e-05,
+      "loss": 1.3694,
+      "step": 16686
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8043574094772339,
+      "learning_rate": 4.559159397303727e-05,
+      "loss": 1.3446,
+      "step": 16687
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.044039487838745,
+      "learning_rate": 4.559132963256675e-05,
+      "loss": 1.4851,
+      "step": 16688
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9048185348510742,
+      "learning_rate": 4.5591065292096224e-05,
+      "loss": 1.3344,
+      "step": 16689
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.04573392868042,
+      "learning_rate": 4.55908009516257e-05,
+      "loss": 1.4126,
+      "step": 16690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7671701908111572,
+      "learning_rate": 4.559053661115517e-05,
+      "loss": 1.4068,
+      "step": 16691
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.426854372024536,
+      "learning_rate": 4.5590272270684645e-05,
+      "loss": 1.4969,
+      "step": 16692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.333129405975342,
+      "learning_rate": 4.559000793021412e-05,
+      "loss": 1.6444,
+      "step": 16693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6712660789489746,
+      "learning_rate": 4.558974358974359e-05,
+      "loss": 1.5505,
+      "step": 16694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7034902572631836,
+      "learning_rate": 4.5589479249273066e-05,
+      "loss": 1.353,
+      "step": 16695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2906851768493652,
+      "learning_rate": 4.558921490880254e-05,
+      "loss": 1.3417,
+      "step": 16696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0996510982513428,
+      "learning_rate": 4.558895056833201e-05,
+      "loss": 1.4481,
+      "step": 16697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.093907117843628,
+      "learning_rate": 4.5588686227861486e-05,
+      "loss": 1.6192,
+      "step": 16698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.638833999633789,
+      "learning_rate": 4.5588421887390967e-05,
+      "loss": 1.3918,
+      "step": 16699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8509337902069092,
+      "learning_rate": 4.5588157546920433e-05,
+      "loss": 1.3749,
+      "step": 16700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7804228067398071,
+      "learning_rate": 4.5587893206449914e-05,
+      "loss": 1.4358,
+      "step": 16701
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9035191535949707,
+      "learning_rate": 4.558762886597938e-05,
+      "loss": 1.2923,
+      "step": 16702
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.548000693321228,
+      "learning_rate": 4.558736452550886e-05,
+      "loss": 1.3435,
+      "step": 16703
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6973576545715332,
+      "learning_rate": 4.5587100185038334e-05,
+      "loss": 1.2631,
+      "step": 16704
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9044655561447144,
+      "learning_rate": 4.55868358445678e-05,
+      "loss": 1.4114,
+      "step": 16705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3667436838150024,
+      "learning_rate": 4.558657150409728e-05,
+      "loss": 1.8469,
+      "step": 16706
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0629072189331055,
+      "learning_rate": 4.558630716362675e-05,
+      "loss": 1.4101,
+      "step": 16707
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6007407903671265,
+      "learning_rate": 4.558604282315623e-05,
+      "loss": 1.2714,
+      "step": 16708
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6334079504013062,
+      "learning_rate": 4.55857784826857e-05,
+      "loss": 1.4161,
+      "step": 16709
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5368709564208984,
+      "learning_rate": 4.5585514142215176e-05,
+      "loss": 1.2934,
+      "step": 16710
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4117649793624878,
+      "learning_rate": 4.558524980174465e-05,
+      "loss": 1.3862,
+      "step": 16711
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.252092957496643,
+      "learning_rate": 4.558498546127412e-05,
+      "loss": 1.1911,
+      "step": 16712
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5145233869552612,
+      "learning_rate": 4.5584721120803597e-05,
+      "loss": 1.3434,
+      "step": 16713
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5403378009796143,
+      "learning_rate": 4.558445678033307e-05,
+      "loss": 1.412,
+      "step": 16714
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6970620155334473,
+      "learning_rate": 4.5584192439862544e-05,
+      "loss": 1.3851,
+      "step": 16715
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0527164936065674,
+      "learning_rate": 4.558392809939202e-05,
+      "loss": 1.4728,
+      "step": 16716
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.876784324645996,
+      "learning_rate": 4.558366375892149e-05,
+      "loss": 1.4665,
+      "step": 16717
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8445093631744385,
+      "learning_rate": 4.5583399418450964e-05,
+      "loss": 1.3219,
+      "step": 16718
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5613795518875122,
+      "learning_rate": 4.5583135077980445e-05,
+      "loss": 1.1631,
+      "step": 16719
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7245104312896729,
+      "learning_rate": 4.558287073750991e-05,
+      "loss": 1.2743,
+      "step": 16720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7580205202102661,
+      "learning_rate": 4.558260639703939e-05,
+      "loss": 1.3393,
+      "step": 16721
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9028584957122803,
+      "learning_rate": 4.558234205656886e-05,
+      "loss": 1.4507,
+      "step": 16722
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.59281587600708,
+      "learning_rate": 4.558207771609834e-05,
+      "loss": 1.4388,
+      "step": 16723
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3801658153533936,
+      "learning_rate": 4.558181337562781e-05,
+      "loss": 1.4931,
+      "step": 16724
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9675308465957642,
+      "learning_rate": 4.5581549035157286e-05,
+      "loss": 1.3497,
+      "step": 16725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.850201964378357,
+      "learning_rate": 4.558128469468676e-05,
+      "loss": 1.7884,
+      "step": 16726
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.942341685295105,
+      "learning_rate": 4.558102035421623e-05,
+      "loss": 1.2738,
+      "step": 16727
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8614776134490967,
+      "learning_rate": 4.558075601374571e-05,
+      "loss": 1.5281,
+      "step": 16728
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.599546194076538,
+      "learning_rate": 4.558049167327518e-05,
+      "loss": 1.4979,
+      "step": 16729
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.733927607536316,
+      "learning_rate": 4.5580227332804654e-05,
+      "loss": 1.3718,
+      "step": 16730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7301015853881836,
+      "learning_rate": 4.557996299233413e-05,
+      "loss": 1.3287,
+      "step": 16731
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.133565902709961,
+      "learning_rate": 4.55796986518636e-05,
+      "loss": 1.4083,
+      "step": 16732
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6892179250717163,
+      "learning_rate": 4.5579434311393075e-05,
+      "loss": 1.3628,
+      "step": 16733
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.976898431777954,
+      "learning_rate": 4.5579169970922555e-05,
+      "loss": 1.4224,
+      "step": 16734
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8790870904922485,
+      "learning_rate": 4.557890563045202e-05,
+      "loss": 1.261,
+      "step": 16735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6139546632766724,
+      "learning_rate": 4.55786412899815e-05,
+      "loss": 1.4578,
+      "step": 16736
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7370893955230713,
+      "learning_rate": 4.557837694951097e-05,
+      "loss": 1.3735,
+      "step": 16737
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7932182550430298,
+      "learning_rate": 4.557811260904045e-05,
+      "loss": 1.2328,
+      "step": 16738
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1661200523376465,
+      "learning_rate": 4.557784826856992e-05,
+      "loss": 1.53,
+      "step": 16739
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7729355096817017,
+      "learning_rate": 4.5577583928099397e-05,
+      "loss": 1.3103,
+      "step": 16740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5075743198394775,
+      "learning_rate": 4.557731958762887e-05,
+      "loss": 1.6449,
+      "step": 16741
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9889936447143555,
+      "learning_rate": 4.557705524715834e-05,
+      "loss": 1.2037,
+      "step": 16742
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9977538585662842,
+      "learning_rate": 4.557679090668782e-05,
+      "loss": 1.3941,
+      "step": 16743
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.526313543319702,
+      "learning_rate": 4.557652656621729e-05,
+      "loss": 1.4225,
+      "step": 16744
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.005815029144287,
+      "learning_rate": 4.5576262225746764e-05,
+      "loss": 1.3436,
+      "step": 16745
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.366830825805664,
+      "learning_rate": 4.557599788527624e-05,
+      "loss": 1.2956,
+      "step": 16746
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2864503860473633,
+      "learning_rate": 4.557573354480571e-05,
+      "loss": 1.5971,
+      "step": 16747
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8343415260314941,
+      "learning_rate": 4.5575469204335185e-05,
+      "loss": 1.4391,
+      "step": 16748
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.803549885749817,
+      "learning_rate": 4.5575204863864665e-05,
+      "loss": 1.5099,
+      "step": 16749
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8833909034729004,
+      "learning_rate": 4.557494052339413e-05,
+      "loss": 1.2433,
+      "step": 16750
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.53094220161438,
+      "learning_rate": 4.5574676182923606e-05,
+      "loss": 1.5161,
+      "step": 16751
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6608176231384277,
+      "learning_rate": 4.557441184245308e-05,
+      "loss": 1.5198,
+      "step": 16752
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.021538734436035,
+      "learning_rate": 4.557414750198255e-05,
+      "loss": 1.3214,
+      "step": 16753
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.389509916305542,
+      "learning_rate": 4.557388316151203e-05,
+      "loss": 1.4917,
+      "step": 16754
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.950400948524475,
+      "learning_rate": 4.55736188210415e-05,
+      "loss": 1.4006,
+      "step": 16755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8441815376281738,
+      "learning_rate": 4.557335448057098e-05,
+      "loss": 1.4639,
+      "step": 16756
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7866698503494263,
+      "learning_rate": 4.557309014010045e-05,
+      "loss": 1.3627,
+      "step": 16757
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.486158847808838,
+      "learning_rate": 4.557282579962993e-05,
+      "loss": 1.4659,
+      "step": 16758
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9736157655715942,
+      "learning_rate": 4.55725614591594e-05,
+      "loss": 1.4289,
+      "step": 16759
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9947924613952637,
+      "learning_rate": 4.5572297118688875e-05,
+      "loss": 1.4177,
+      "step": 16760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6389647722244263,
+      "learning_rate": 4.557203277821835e-05,
+      "loss": 1.4949,
+      "step": 16761
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.738218069076538,
+      "learning_rate": 4.557176843774782e-05,
+      "loss": 1.2793,
+      "step": 16762
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.205000162124634,
+      "learning_rate": 4.5571504097277295e-05,
+      "loss": 2.858,
+      "step": 16763
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.811723232269287,
+      "learning_rate": 4.557123975680677e-05,
+      "loss": 1.8801,
+      "step": 16764
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0159075260162354,
+      "learning_rate": 4.557097541633624e-05,
+      "loss": 1.3753,
+      "step": 16765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7881535291671753,
+      "learning_rate": 4.5570711075865716e-05,
+      "loss": 1.4676,
+      "step": 16766
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.521118402481079,
+      "learning_rate": 4.557044673539519e-05,
+      "loss": 1.4794,
+      "step": 16767
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8077054023742676,
+      "learning_rate": 4.557018239492466e-05,
+      "loss": 1.3507,
+      "step": 16768
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9505085945129395,
+      "learning_rate": 4.5569918054454144e-05,
+      "loss": 1.3077,
+      "step": 16769
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8381097316741943,
+      "learning_rate": 4.556965371398361e-05,
+      "loss": 1.3729,
+      "step": 16770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6528964042663574,
+      "learning_rate": 4.556938937351309e-05,
+      "loss": 1.5238,
+      "step": 16771
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6807304620742798,
+      "learning_rate": 4.556912503304256e-05,
+      "loss": 1.3741,
+      "step": 16772
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7750065326690674,
+      "learning_rate": 4.556886069257204e-05,
+      "loss": 1.5002,
+      "step": 16773
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.71102774143219,
+      "learning_rate": 4.556859635210151e-05,
+      "loss": 1.3655,
+      "step": 16774
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2666831016540527,
+      "learning_rate": 4.5568332011630985e-05,
+      "loss": 1.33,
+      "step": 16775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1554512977600098,
+      "learning_rate": 4.556806767116046e-05,
+      "loss": 1.3891,
+      "step": 16776
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.705659031867981,
+      "learning_rate": 4.5567803330689925e-05,
+      "loss": 1.3131,
+      "step": 16777
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6191266775131226,
+      "learning_rate": 4.5567538990219406e-05,
+      "loss": 1.3485,
+      "step": 16778
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7665772438049316,
+      "learning_rate": 4.556727464974888e-05,
+      "loss": 1.2668,
+      "step": 16779
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0034568309783936,
+      "learning_rate": 4.556701030927835e-05,
+      "loss": 1.3541,
+      "step": 16780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.172210931777954,
+      "learning_rate": 4.5566745968807827e-05,
+      "loss": 1.3402,
+      "step": 16781
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6763246059417725,
+      "learning_rate": 4.55664816283373e-05,
+      "loss": 1.2395,
+      "step": 16782
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5101900100708008,
+      "learning_rate": 4.5566217287866774e-05,
+      "loss": 1.2725,
+      "step": 16783
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7804616689682007,
+      "learning_rate": 4.5565952947396254e-05,
+      "loss": 1.2888,
+      "step": 16784
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5776302814483643,
+      "learning_rate": 4.556568860692572e-05,
+      "loss": 1.6578,
+      "step": 16785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0460171699523926,
+      "learning_rate": 4.55654242664552e-05,
+      "loss": 1.5964,
+      "step": 16786
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6370301246643066,
+      "learning_rate": 4.556515992598467e-05,
+      "loss": 1.3156,
+      "step": 16787
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2254860401153564,
+      "learning_rate": 4.556489558551414e-05,
+      "loss": 1.5813,
+      "step": 16788
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7151700258255005,
+      "learning_rate": 4.556463124504362e-05,
+      "loss": 2.556,
+      "step": 16789
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6881204843521118,
+      "learning_rate": 4.556436690457309e-05,
+      "loss": 1.4271,
+      "step": 16790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1550989151000977,
+      "learning_rate": 4.556410256410257e-05,
+      "loss": 1.4262,
+      "step": 16791
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9804729223251343,
+      "learning_rate": 4.5563838223632036e-05,
+      "loss": 1.5033,
+      "step": 16792
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6214380264282227,
+      "learning_rate": 4.5563573883161516e-05,
+      "loss": 1.4173,
+      "step": 16793
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5442848205566406,
+      "learning_rate": 4.556330954269099e-05,
+      "loss": 1.3613,
+      "step": 16794
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1819822788238525,
+      "learning_rate": 4.556304520222046e-05,
+      "loss": 1.5194,
+      "step": 16795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9372062683105469,
+      "learning_rate": 4.556278086174994e-05,
+      "loss": 1.3702,
+      "step": 16796
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.280914545059204,
+      "learning_rate": 4.556251652127941e-05,
+      "loss": 1.4256,
+      "step": 16797
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7776607275009155,
+      "learning_rate": 4.5562252180808884e-05,
+      "loss": 1.3384,
+      "step": 16798
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5956909656524658,
+      "learning_rate": 4.556198784033836e-05,
+      "loss": 1.2675,
+      "step": 16799
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6216496229171753,
+      "learning_rate": 4.556172349986783e-05,
+      "loss": 1.2891,
+      "step": 16800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4193575382232666,
+      "learning_rate": 4.5561459159397305e-05,
+      "loss": 1.6027,
+      "step": 16801
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7293182611465454,
+      "learning_rate": 4.556119481892678e-05,
+      "loss": 1.3017,
+      "step": 16802
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.002964735031128,
+      "learning_rate": 4.556093047845625e-05,
+      "loss": 1.4505,
+      "step": 16803
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.065373659133911,
+      "learning_rate": 4.556066613798573e-05,
+      "loss": 1.3863,
+      "step": 16804
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5696072578430176,
+      "learning_rate": 4.55604017975152e-05,
+      "loss": 1.3167,
+      "step": 16805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7826721668243408,
+      "learning_rate": 4.556013745704468e-05,
+      "loss": 1.4553,
+      "step": 16806
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6086653470993042,
+      "learning_rate": 4.5559873116574146e-05,
+      "loss": 1.4187,
+      "step": 16807
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8959293365478516,
+      "learning_rate": 4.5559608776103626e-05,
+      "loss": 1.4204,
+      "step": 16808
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0795578956604004,
+      "learning_rate": 4.55593444356331e-05,
+      "loss": 1.4238,
+      "step": 16809
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.39524245262146,
+      "learning_rate": 4.5559080095162574e-05,
+      "loss": 1.4782,
+      "step": 16810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4892135858535767,
+      "learning_rate": 4.555881575469205e-05,
+      "loss": 1.3137,
+      "step": 16811
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6563116312026978,
+      "learning_rate": 4.555855141422152e-05,
+      "loss": 1.4552,
+      "step": 16812
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5578937530517578,
+      "learning_rate": 4.5558287073750994e-05,
+      "loss": 1.3217,
+      "step": 16813
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1547861099243164,
+      "learning_rate": 4.555802273328047e-05,
+      "loss": 1.487,
+      "step": 16814
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.067795753479004,
+      "learning_rate": 4.555775839280994e-05,
+      "loss": 1.6982,
+      "step": 16815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7368742227554321,
+      "learning_rate": 4.5557494052339415e-05,
+      "loss": 1.3094,
+      "step": 16816
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.799916386604309,
+      "learning_rate": 4.555722971186889e-05,
+      "loss": 1.8365,
+      "step": 16817
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.036500930786133,
+      "learning_rate": 4.555696537139836e-05,
+      "loss": 1.3421,
+      "step": 16818
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.709369421005249,
+      "learning_rate": 4.555670103092784e-05,
+      "loss": 2.4508,
+      "step": 16819
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8814371824264526,
+      "learning_rate": 4.555643669045731e-05,
+      "loss": 1.3885,
+      "step": 16820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6307969093322754,
+      "learning_rate": 4.555617234998679e-05,
+      "loss": 1.306,
+      "step": 16821
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5337027311325073,
+      "learning_rate": 4.5555908009516256e-05,
+      "loss": 1.26,
+      "step": 16822
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5239506959915161,
+      "learning_rate": 4.555564366904573e-05,
+      "loss": 1.5167,
+      "step": 16823
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.945367455482483,
+      "learning_rate": 4.555537932857521e-05,
+      "loss": 1.3022,
+      "step": 16824
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4768486022949219,
+      "learning_rate": 4.555511498810468e-05,
+      "loss": 1.2572,
+      "step": 16825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4503909349441528,
+      "learning_rate": 4.555485064763416e-05,
+      "loss": 1.2682,
+      "step": 16826
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.726394534111023,
+      "learning_rate": 4.5554586307163624e-05,
+      "loss": 1.425,
+      "step": 16827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6155717372894287,
+      "learning_rate": 4.5554321966693105e-05,
+      "loss": 1.2826,
+      "step": 16828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.675521969795227,
+      "learning_rate": 4.555405762622258e-05,
+      "loss": 1.4412,
+      "step": 16829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1667134761810303,
+      "learning_rate": 4.555379328575205e-05,
+      "loss": 1.5663,
+      "step": 16830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5831947326660156,
+      "learning_rate": 4.5553528945281525e-05,
+      "loss": 2.721,
+      "step": 16831
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4292495250701904,
+      "learning_rate": 4.5553264604811e-05,
+      "loss": 1.512,
+      "step": 16832
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3901751041412354,
+      "learning_rate": 4.555300026434047e-05,
+      "loss": 1.2123,
+      "step": 16833
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.917894959449768,
+      "learning_rate": 4.5552735923869946e-05,
+      "loss": 1.4352,
+      "step": 16834
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5396993160247803,
+      "learning_rate": 4.555247158339942e-05,
+      "loss": 1.3143,
+      "step": 16835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4619863033294678,
+      "learning_rate": 4.555220724292889e-05,
+      "loss": 1.6186,
+      "step": 16836
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5084539651870728,
+      "learning_rate": 4.555194290245837e-05,
+      "loss": 1.3363,
+      "step": 16837
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.090210199356079,
+      "learning_rate": 4.555167856198784e-05,
+      "loss": 1.4592,
+      "step": 16838
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.581272840499878,
+      "learning_rate": 4.555141422151732e-05,
+      "loss": 1.3292,
+      "step": 16839
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2354495525360107,
+      "learning_rate": 4.555114988104679e-05,
+      "loss": 1.4193,
+      "step": 16840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.040419340133667,
+      "learning_rate": 4.555088554057627e-05,
+      "loss": 1.2546,
+      "step": 16841
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9151427745819092,
+      "learning_rate": 4.5550621200105735e-05,
+      "loss": 1.691,
+      "step": 16842
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.138094425201416,
+      "learning_rate": 4.5550356859635215e-05,
+      "loss": 1.5013,
+      "step": 16843
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7928115129470825,
+      "learning_rate": 4.555009251916469e-05,
+      "loss": 1.4414,
+      "step": 16844
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0980775356292725,
+      "learning_rate": 4.554982817869416e-05,
+      "loss": 1.3949,
+      "step": 16845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9780356884002686,
+      "learning_rate": 4.5549563838223636e-05,
+      "loss": 1.5249,
+      "step": 16846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5300424098968506,
+      "learning_rate": 4.554929949775311e-05,
+      "loss": 1.2654,
+      "step": 16847
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9665682315826416,
+      "learning_rate": 4.554903515728258e-05,
+      "loss": 1.423,
+      "step": 16848
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.334277868270874,
+      "learning_rate": 4.5548770816812056e-05,
+      "loss": 1.5052,
+      "step": 16849
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4348323345184326,
+      "learning_rate": 4.554850647634153e-05,
+      "loss": 1.3685,
+      "step": 16850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0435190200805664,
+      "learning_rate": 4.5548242135871004e-05,
+      "loss": 1.4129,
+      "step": 16851
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.551259994506836,
+      "learning_rate": 4.554797779540048e-05,
+      "loss": 1.4251,
+      "step": 16852
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.658578872680664,
+      "learning_rate": 4.554771345492995e-05,
+      "loss": 1.432,
+      "step": 16853
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.556917428970337,
+      "learning_rate": 4.554744911445943e-05,
+      "loss": 1.2413,
+      "step": 16854
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6530201435089111,
+      "learning_rate": 4.55471847739889e-05,
+      "loss": 1.3686,
+      "step": 16855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.492152452468872,
+      "learning_rate": 4.554692043351838e-05,
+      "loss": 1.4509,
+      "step": 16856
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.614902377128601,
+      "learning_rate": 4.5546656093047845e-05,
+      "loss": 1.3308,
+      "step": 16857
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8058955669403076,
+      "learning_rate": 4.5546391752577325e-05,
+      "loss": 1.3929,
+      "step": 16858
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.392647385597229,
+      "learning_rate": 4.55461274121068e-05,
+      "loss": 1.3055,
+      "step": 16859
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8003530502319336,
+      "learning_rate": 4.5545863071636266e-05,
+      "loss": 1.4382,
+      "step": 16860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.197916030883789,
+      "learning_rate": 4.5545598731165746e-05,
+      "loss": 1.4617,
+      "step": 16861
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9077911376953125,
+      "learning_rate": 4.554533439069521e-05,
+      "loss": 1.441,
+      "step": 16862
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4628396034240723,
+      "learning_rate": 4.554507005022469e-05,
+      "loss": 1.3459,
+      "step": 16863
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5677831172943115,
+      "learning_rate": 4.554480570975417e-05,
+      "loss": 1.3207,
+      "step": 16864
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9955170154571533,
+      "learning_rate": 4.554454136928364e-05,
+      "loss": 1.4381,
+      "step": 16865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3534742593765259,
+      "learning_rate": 4.5544277028813114e-05,
+      "loss": 1.3062,
+      "step": 16866
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4022598266601562,
+      "learning_rate": 4.554401268834259e-05,
+      "loss": 2.2253,
+      "step": 16867
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.283095359802246,
+      "learning_rate": 4.554374834787206e-05,
+      "loss": 1.2655,
+      "step": 16868
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3394124507904053,
+      "learning_rate": 4.5543484007401535e-05,
+      "loss": 2.2984,
+      "step": 16869
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7381980419158936,
+      "learning_rate": 4.554321966693101e-05,
+      "loss": 1.2129,
+      "step": 16870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.844188928604126,
+      "learning_rate": 4.554295532646048e-05,
+      "loss": 1.3636,
+      "step": 16871
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8281687498092651,
+      "learning_rate": 4.5542690985989955e-05,
+      "loss": 1.3255,
+      "step": 16872
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0584726333618164,
+      "learning_rate": 4.554242664551943e-05,
+      "loss": 1.4085,
+      "step": 16873
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8861666917800903,
+      "learning_rate": 4.554216230504891e-05,
+      "loss": 1.3773,
+      "step": 16874
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2670464515686035,
+      "learning_rate": 4.5541897964578376e-05,
+      "loss": 1.392,
+      "step": 16875
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3656057119369507,
+      "learning_rate": 4.5541633624107856e-05,
+      "loss": 1.6218,
+      "step": 16876
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9849926233291626,
+      "learning_rate": 4.554136928363732e-05,
+      "loss": 1.5135,
+      "step": 16877
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7166517972946167,
+      "learning_rate": 4.5541104943166804e-05,
+      "loss": 1.3713,
+      "step": 16878
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7854394912719727,
+      "learning_rate": 4.554084060269628e-05,
+      "loss": 1.5312,
+      "step": 16879
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7005040645599365,
+      "learning_rate": 4.554057626222575e-05,
+      "loss": 1.384,
+      "step": 16880
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.431246042251587,
+      "learning_rate": 4.5540311921755224e-05,
+      "loss": 1.2939,
+      "step": 16881
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.179642915725708,
+      "learning_rate": 4.55400475812847e-05,
+      "loss": 1.4493,
+      "step": 16882
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.533942461013794,
+      "learning_rate": 4.553978324081417e-05,
+      "loss": 1.6707,
+      "step": 16883
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8474332094192505,
+      "learning_rate": 4.5539518900343645e-05,
+      "loss": 1.5068,
+      "step": 16884
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0132884979248047,
+      "learning_rate": 4.553925455987312e-05,
+      "loss": 1.4433,
+      "step": 16885
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.976222038269043,
+      "learning_rate": 4.553899021940259e-05,
+      "loss": 1.3998,
+      "step": 16886
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.106632947921753,
+      "learning_rate": 4.5538725878932066e-05,
+      "loss": 1.3525,
+      "step": 16887
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9756457805633545,
+      "learning_rate": 4.553846153846154e-05,
+      "loss": 1.3468,
+      "step": 16888
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6872920989990234,
+      "learning_rate": 4.553819719799102e-05,
+      "loss": 1.3611,
+      "step": 16889
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.898322582244873,
+      "learning_rate": 4.5537932857520486e-05,
+      "loss": 1.45,
+      "step": 16890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4374094009399414,
+      "learning_rate": 4.553766851704997e-05,
+      "loss": 1.263,
+      "step": 16891
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5307260751724243,
+      "learning_rate": 4.5537404176579434e-05,
+      "loss": 1.334,
+      "step": 16892
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4463820457458496,
+      "learning_rate": 4.5537139836108914e-05,
+      "loss": 1.4605,
+      "step": 16893
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4064419269561768,
+      "learning_rate": 4.553687549563839e-05,
+      "loss": 1.4976,
+      "step": 16894
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.595046043395996,
+      "learning_rate": 4.5536611155167854e-05,
+      "loss": 1.3199,
+      "step": 16895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.221104145050049,
+      "learning_rate": 4.5536346814697335e-05,
+      "loss": 1.5226,
+      "step": 16896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8224313259124756,
+      "learning_rate": 4.55360824742268e-05,
+      "loss": 1.3546,
+      "step": 16897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0992090702056885,
+      "learning_rate": 4.553581813375628e-05,
+      "loss": 1.5963,
+      "step": 16898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2122914791107178,
+      "learning_rate": 4.5535553793285755e-05,
+      "loss": 1.0852,
+      "step": 16899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8735954761505127,
+      "learning_rate": 4.553528945281523e-05,
+      "loss": 1.5419,
+      "step": 16900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0833206176757812,
+      "learning_rate": 4.55350251123447e-05,
+      "loss": 1.3565,
+      "step": 16901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8098682165145874,
+      "learning_rate": 4.5534760771874176e-05,
+      "loss": 1.2568,
+      "step": 16902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.174478530883789,
+      "learning_rate": 4.553449643140365e-05,
+      "loss": 1.4839,
+      "step": 16903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8012086153030396,
+      "learning_rate": 4.553423209093313e-05,
+      "loss": 1.3191,
+      "step": 16904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9998160600662231,
+      "learning_rate": 4.55339677504626e-05,
+      "loss": 1.3447,
+      "step": 16905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4608309268951416,
+      "learning_rate": 4.553370340999207e-05,
+      "loss": 1.3819,
+      "step": 16906
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.720167875289917,
+      "learning_rate": 4.5533439069521544e-05,
+      "loss": 1.6791,
+      "step": 16907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6846832036972046,
+      "learning_rate": 4.553317472905102e-05,
+      "loss": 1.4791,
+      "step": 16908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7264515161514282,
+      "learning_rate": 4.55329103885805e-05,
+      "loss": 1.298,
+      "step": 16909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6786086559295654,
+      "learning_rate": 4.5532646048109965e-05,
+      "loss": 1.3238,
+      "step": 16910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7297379970550537,
+      "learning_rate": 4.5532381707639445e-05,
+      "loss": 1.5308,
+      "step": 16911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5425167083740234,
+      "learning_rate": 4.553211736716891e-05,
+      "loss": 1.3598,
+      "step": 16912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1934995651245117,
+      "learning_rate": 4.553185302669839e-05,
+      "loss": 1.42,
+      "step": 16913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5456360578536987,
+      "learning_rate": 4.5531588686227866e-05,
+      "loss": 1.3338,
+      "step": 16914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3656834363937378,
+      "learning_rate": 4.553132434575734e-05,
+      "loss": 1.3031,
+      "step": 16915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4606086015701294,
+      "learning_rate": 4.553106000528681e-05,
+      "loss": 1.6895,
+      "step": 16916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6305129528045654,
+      "learning_rate": 4.5530795664816286e-05,
+      "loss": 1.326,
+      "step": 16917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.96783447265625,
+      "learning_rate": 4.553053132434576e-05,
+      "loss": 1.3872,
+      "step": 16918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.425904631614685,
+      "learning_rate": 4.5530266983875234e-05,
+      "loss": 1.3304,
+      "step": 16919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5325639247894287,
+      "learning_rate": 4.553000264340471e-05,
+      "loss": 1.2233,
+      "step": 16920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3032631874084473,
+      "learning_rate": 4.552973830293418e-05,
+      "loss": 1.3497,
+      "step": 16921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.529555082321167,
+      "learning_rate": 4.5529473962463654e-05,
+      "loss": 1.2693,
+      "step": 16922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.498497724533081,
+      "learning_rate": 4.552920962199313e-05,
+      "loss": 1.4796,
+      "step": 16923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.653370976448059,
+      "learning_rate": 4.552894528152261e-05,
+      "loss": 1.4479,
+      "step": 16924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6547846794128418,
+      "learning_rate": 4.5528680941052075e-05,
+      "loss": 1.3015,
+      "step": 16925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7237327098846436,
+      "learning_rate": 4.5528416600581555e-05,
+      "loss": 1.5293,
+      "step": 16926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5501348972320557,
+      "learning_rate": 4.552815226011102e-05,
+      "loss": 1.4135,
+      "step": 16927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.152034282684326,
+      "learning_rate": 4.55278879196405e-05,
+      "loss": 1.2584,
+      "step": 16928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7224655151367188,
+      "learning_rate": 4.5527623579169976e-05,
+      "loss": 1.4107,
+      "step": 16929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0632076263427734,
+      "learning_rate": 4.552735923869945e-05,
+      "loss": 1.469,
+      "step": 16930
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2842917442321777,
+      "learning_rate": 4.552709489822892e-05,
+      "loss": 1.4184,
+      "step": 16931
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.433044672012329,
+      "learning_rate": 4.552683055775839e-05,
+      "loss": 1.2444,
+      "step": 16932
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.732826828956604,
+      "learning_rate": 4.552656621728787e-05,
+      "loss": 2.3825,
+      "step": 16933
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.251716375350952,
+      "learning_rate": 4.5526301876817344e-05,
+      "loss": 1.2861,
+      "step": 16934
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5974957942962646,
+      "learning_rate": 4.552603753634682e-05,
+      "loss": 1.4734,
+      "step": 16935
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5403467416763306,
+      "learning_rate": 4.552577319587629e-05,
+      "loss": 1.296,
+      "step": 16936
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8232958316802979,
+      "learning_rate": 4.5525508855405765e-05,
+      "loss": 1.3415,
+      "step": 16937
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8487178087234497,
+      "learning_rate": 4.552524451493524e-05,
+      "loss": 1.3362,
+      "step": 16938
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.010042905807495,
+      "learning_rate": 4.552498017446472e-05,
+      "loss": 2.3517,
+      "step": 16939
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5865066051483154,
+      "learning_rate": 4.5524715833994185e-05,
+      "loss": 1.4265,
+      "step": 16940
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1944289207458496,
+      "learning_rate": 4.552445149352366e-05,
+      "loss": 1.4685,
+      "step": 16941
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6564923524856567,
+      "learning_rate": 4.552418715305313e-05,
+      "loss": 1.3642,
+      "step": 16942
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9817910194396973,
+      "learning_rate": 4.5523922812582606e-05,
+      "loss": 1.5084,
+      "step": 16943
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.709083080291748,
+      "learning_rate": 4.5523658472112086e-05,
+      "loss": 1.4294,
+      "step": 16944
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6587707996368408,
+      "learning_rate": 4.552339413164155e-05,
+      "loss": 1.3223,
+      "step": 16945
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3995587825775146,
+      "learning_rate": 4.5523129791171033e-05,
+      "loss": 1.4348,
+      "step": 16946
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.970351219177246,
+      "learning_rate": 4.55228654507005e-05,
+      "loss": 1.2781,
+      "step": 16947
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8033591508865356,
+      "learning_rate": 4.552260111022998e-05,
+      "loss": 1.3817,
+      "step": 16948
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0507736206054688,
+      "learning_rate": 4.5522336769759454e-05,
+      "loss": 1.4961,
+      "step": 16949
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0111539363861084,
+      "learning_rate": 4.552207242928893e-05,
+      "loss": 1.5256,
+      "step": 16950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9066709280014038,
+      "learning_rate": 4.55218080888184e-05,
+      "loss": 1.4192,
+      "step": 16951
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7816319465637207,
+      "learning_rate": 4.5521543748347875e-05,
+      "loss": 1.4006,
+      "step": 16952
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8772767782211304,
+      "learning_rate": 4.552127940787735e-05,
+      "loss": 1.4323,
+      "step": 16953
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9375905990600586,
+      "learning_rate": 4.552101506740682e-05,
+      "loss": 1.3361,
+      "step": 16954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6783558130264282,
+      "learning_rate": 4.5520750726936296e-05,
+      "loss": 1.3167,
+      "step": 16955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0394725799560547,
+      "learning_rate": 4.552048638646577e-05,
+      "loss": 1.4494,
+      "step": 16956
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5317797660827637,
+      "learning_rate": 4.552022204599524e-05,
+      "loss": 1.4561,
+      "step": 16957
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6301941871643066,
+      "learning_rate": 4.5519957705524716e-05,
+      "loss": 1.513,
+      "step": 16958
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7244693040847778,
+      "learning_rate": 4.55196933650542e-05,
+      "loss": 1.2815,
+      "step": 16959
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.500184416770935,
+      "learning_rate": 4.5519429024583663e-05,
+      "loss": 1.3007,
+      "step": 16960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1609835624694824,
+      "learning_rate": 4.5519164684113144e-05,
+      "loss": 1.3946,
+      "step": 16961
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7053757905960083,
+      "learning_rate": 4.551890034364261e-05,
+      "loss": 1.4139,
+      "step": 16962
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0786705017089844,
+      "learning_rate": 4.551863600317209e-05,
+      "loss": 1.4382,
+      "step": 16963
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4681153297424316,
+      "learning_rate": 4.5518371662701565e-05,
+      "loss": 1.3276,
+      "step": 16964
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6518621444702148,
+      "learning_rate": 4.551810732223104e-05,
+      "loss": 1.2875,
+      "step": 16965
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6165218353271484,
+      "learning_rate": 4.551784298176051e-05,
+      "loss": 1.3731,
+      "step": 16966
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5992122888565063,
+      "learning_rate": 4.551757864128998e-05,
+      "loss": 1.3375,
+      "step": 16967
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5530238151550293,
+      "learning_rate": 4.551731430081946e-05,
+      "loss": 1.4151,
+      "step": 16968
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0386946201324463,
+      "learning_rate": 4.551704996034893e-05,
+      "loss": 1.8131,
+      "step": 16969
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7173426151275635,
+      "learning_rate": 4.5516785619878406e-05,
+      "loss": 1.506,
+      "step": 16970
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3684825897216797,
+      "learning_rate": 4.551652127940788e-05,
+      "loss": 1.5331,
+      "step": 16971
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7303082942962646,
+      "learning_rate": 4.551625693893735e-05,
+      "loss": 1.5856,
+      "step": 16972
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6768105030059814,
+      "learning_rate": 4.551599259846683e-05,
+      "loss": 1.3406,
+      "step": 16973
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8979735374450684,
+      "learning_rate": 4.551572825799631e-05,
+      "loss": 1.4318,
+      "step": 16974
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8565629720687866,
+      "learning_rate": 4.5515463917525774e-05,
+      "loss": 1.4407,
+      "step": 16975
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.847249984741211,
+      "learning_rate": 4.5515199577055254e-05,
+      "loss": 1.4526,
+      "step": 16976
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.056180238723755,
+      "learning_rate": 4.551493523658472e-05,
+      "loss": 1.299,
+      "step": 16977
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4519671201705933,
+      "learning_rate": 4.5514670896114195e-05,
+      "loss": 1.8717,
+      "step": 16978
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8535724878311157,
+      "learning_rate": 4.5514406555643675e-05,
+      "loss": 1.3178,
+      "step": 16979
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4611423015594482,
+      "learning_rate": 4.551414221517314e-05,
+      "loss": 1.4532,
+      "step": 16980
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1917550563812256,
+      "learning_rate": 4.551387787470262e-05,
+      "loss": 1.4086,
+      "step": 16981
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.828898549079895,
+      "learning_rate": 4.551361353423209e-05,
+      "loss": 1.3935,
+      "step": 16982
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5899006128311157,
+      "learning_rate": 4.551334919376157e-05,
+      "loss": 1.4088,
+      "step": 16983
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5285049676895142,
+      "learning_rate": 4.551308485329104e-05,
+      "loss": 1.3065,
+      "step": 16984
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.059431791305542,
+      "learning_rate": 4.5512820512820516e-05,
+      "loss": 1.4693,
+      "step": 16985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0450761318206787,
+      "learning_rate": 4.551255617234999e-05,
+      "loss": 1.3641,
+      "step": 16986
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4853215217590332,
+      "learning_rate": 4.5512291831879463e-05,
+      "loss": 1.4281,
+      "step": 16987
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6539758443832397,
+      "learning_rate": 4.551202749140894e-05,
+      "loss": 1.4281,
+      "step": 16988
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8623883724212646,
+      "learning_rate": 4.551176315093841e-05,
+      "loss": 1.3361,
+      "step": 16989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7569760084152222,
+      "learning_rate": 4.5511498810467884e-05,
+      "loss": 1.3448,
+      "step": 16990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8516819477081299,
+      "learning_rate": 4.551123446999736e-05,
+      "loss": 1.4382,
+      "step": 16991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7529606819152832,
+      "learning_rate": 4.551097012952683e-05,
+      "loss": 2.4592,
+      "step": 16992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1513218879699707,
+      "learning_rate": 4.5510705789056305e-05,
+      "loss": 1.2887,
+      "step": 16993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6498405933380127,
+      "learning_rate": 4.5510441448585785e-05,
+      "loss": 1.4306,
+      "step": 16994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6369340419769287,
+      "learning_rate": 4.551017710811525e-05,
+      "loss": 1.2053,
+      "step": 16995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.597306966781616,
+      "learning_rate": 4.550991276764473e-05,
+      "loss": 1.5953,
+      "step": 16996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.373622179031372,
+      "learning_rate": 4.55096484271742e-05,
+      "loss": 1.6974,
+      "step": 16997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7668119668960571,
+      "learning_rate": 4.550938408670368e-05,
+      "loss": 1.3977,
+      "step": 16998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3258185386657715,
+      "learning_rate": 4.550911974623315e-05,
+      "loss": 2.2865,
+      "step": 16999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0021774768829346,
+      "learning_rate": 4.550885540576263e-05,
+      "loss": 1.3435,
+      "step": 17000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9656556844711304,
+      "learning_rate": 4.55085910652921e-05,
+      "loss": 1.5423,
+      "step": 17001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.967262864112854,
+      "learning_rate": 4.5508326724821574e-05,
+      "loss": 1.3986,
+      "step": 17002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4617905616760254,
+      "learning_rate": 4.550806238435105e-05,
+      "loss": 1.3596,
+      "step": 17003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2692925930023193,
+      "learning_rate": 4.550779804388052e-05,
+      "loss": 1.3534,
+      "step": 17004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2632193565368652,
+      "learning_rate": 4.5507533703409994e-05,
+      "loss": 1.4325,
+      "step": 17005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7838267087936401,
+      "learning_rate": 4.550726936293947e-05,
+      "loss": 1.485,
+      "step": 17006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8045873641967773,
+      "learning_rate": 4.550700502246894e-05,
+      "loss": 1.3713,
+      "step": 17007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6681331396102905,
+      "learning_rate": 4.5506740681998415e-05,
+      "loss": 1.1919,
+      "step": 17008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3307431936264038,
+      "learning_rate": 4.5506476341527896e-05,
+      "loss": 1.184,
+      "step": 17009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5176010131835938,
+      "learning_rate": 4.550621200105736e-05,
+      "loss": 1.548,
+      "step": 17010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3099793195724487,
+      "learning_rate": 4.550594766058684e-05,
+      "loss": 1.3858,
+      "step": 17011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.510878324508667,
+      "learning_rate": 4.550568332011631e-05,
+      "loss": 2.1963,
+      "step": 17012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4298278093338013,
+      "learning_rate": 4.550541897964578e-05,
+      "loss": 1.341,
+      "step": 17013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6393377780914307,
+      "learning_rate": 4.5505154639175263e-05,
+      "loss": 1.2825,
+      "step": 17014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4677776098251343,
+      "learning_rate": 4.550489029870473e-05,
+      "loss": 1.3272,
+      "step": 17015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7276675701141357,
+      "learning_rate": 4.550462595823421e-05,
+      "loss": 1.346,
+      "step": 17016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.926981806755066,
+      "learning_rate": 4.550436161776368e-05,
+      "loss": 1.4059,
+      "step": 17017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7783571481704712,
+      "learning_rate": 4.550409727729316e-05,
+      "loss": 1.4382,
+      "step": 17018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.270758628845215,
+      "learning_rate": 4.550383293682263e-05,
+      "loss": 1.4153,
+      "step": 17019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2011239528656006,
+      "learning_rate": 4.5503568596352105e-05,
+      "loss": 1.5114,
+      "step": 17020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8302178382873535,
+      "learning_rate": 4.550330425588158e-05,
+      "loss": 1.465,
+      "step": 17021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2911320924758911,
+      "learning_rate": 4.550303991541105e-05,
+      "loss": 1.2707,
+      "step": 17022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3163012266159058,
+      "learning_rate": 4.5502775574940526e-05,
+      "loss": 1.1898,
+      "step": 17023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6929243803024292,
+      "learning_rate": 4.550251123447e-05,
+      "loss": 1.4312,
+      "step": 17024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.075566530227661,
+      "learning_rate": 4.550224689399947e-05,
+      "loss": 1.5259,
+      "step": 17025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.705923318862915,
+      "learning_rate": 4.5501982553528946e-05,
+      "loss": 1.3555,
+      "step": 17026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6047519445419312,
+      "learning_rate": 4.550171821305842e-05,
+      "loss": 1.4528,
+      "step": 17027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4012001752853394,
+      "learning_rate": 4.5501453872587893e-05,
+      "loss": 1.365,
+      "step": 17028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6222879886627197,
+      "learning_rate": 4.5501189532117374e-05,
+      "loss": 1.3665,
+      "step": 17029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8734288215637207,
+      "learning_rate": 4.550092519164684e-05,
+      "loss": 1.2709,
+      "step": 17030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9024957418441772,
+      "learning_rate": 4.550066085117632e-05,
+      "loss": 1.2387,
+      "step": 17031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.564820647239685,
+      "learning_rate": 4.550039651070579e-05,
+      "loss": 1.3122,
+      "step": 17032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8633183240890503,
+      "learning_rate": 4.550013217023527e-05,
+      "loss": 1.3446,
+      "step": 17033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7935130596160889,
+      "learning_rate": 4.549986782976474e-05,
+      "loss": 1.3789,
+      "step": 17034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8437614440917969,
+      "learning_rate": 4.5499603489294215e-05,
+      "loss": 1.403,
+      "step": 17035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6411612033843994,
+      "learning_rate": 4.549933914882369e-05,
+      "loss": 1.3737,
+      "step": 17036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7270102500915527,
+      "learning_rate": 4.549907480835316e-05,
+      "loss": 1.5101,
+      "step": 17037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.296260118484497,
+      "learning_rate": 4.5498810467882636e-05,
+      "loss": 1.6155,
+      "step": 17038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9537643194198608,
+      "learning_rate": 4.549854612741211e-05,
+      "loss": 1.3018,
+      "step": 17039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9109117984771729,
+      "learning_rate": 4.549828178694158e-05,
+      "loss": 1.3137,
+      "step": 17040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.527028203010559,
+      "learning_rate": 4.5498017446471057e-05,
+      "loss": 1.2432,
+      "step": 17041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7481951713562012,
+      "learning_rate": 4.549775310600053e-05,
+      "loss": 1.5837,
+      "step": 17042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.769018530845642,
+      "learning_rate": 4.5497488765530004e-05,
+      "loss": 1.5216,
+      "step": 17043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6730026006698608,
+      "learning_rate": 4.5497224425059484e-05,
+      "loss": 1.2559,
+      "step": 17044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7237950563430786,
+      "learning_rate": 4.549696008458895e-05,
+      "loss": 1.5026,
+      "step": 17045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.14486026763916,
+      "learning_rate": 4.549669574411843e-05,
+      "loss": 1.1274,
+      "step": 17046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5769636631011963,
+      "learning_rate": 4.54964314036479e-05,
+      "loss": 1.7585,
+      "step": 17047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7005345821380615,
+      "learning_rate": 4.549616706317738e-05,
+      "loss": 1.2355,
+      "step": 17048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5836334228515625,
+      "learning_rate": 4.549590272270685e-05,
+      "loss": 1.2196,
+      "step": 17049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4687776565551758,
+      "learning_rate": 4.549563838223632e-05,
+      "loss": 1.4439,
+      "step": 17050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.712389349937439,
+      "learning_rate": 4.54953740417658e-05,
+      "loss": 1.3642,
+      "step": 17051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8012948036193848,
+      "learning_rate": 4.5495109701295266e-05,
+      "loss": 1.4436,
+      "step": 17052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6278018951416016,
+      "learning_rate": 4.5494845360824746e-05,
+      "loss": 1.7514,
+      "step": 17053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5045567750930786,
+      "learning_rate": 4.549458102035422e-05,
+      "loss": 1.3485,
+      "step": 17054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.903568148612976,
+      "learning_rate": 4.549431667988369e-05,
+      "loss": 1.3411,
+      "step": 17055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9892336130142212,
+      "learning_rate": 4.549405233941317e-05,
+      "loss": 1.6071,
+      "step": 17056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7112711668014526,
+      "learning_rate": 4.549378799894264e-05,
+      "loss": 1.4697,
+      "step": 17057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9019695520401,
+      "learning_rate": 4.5493523658472114e-05,
+      "loss": 1.4068,
+      "step": 17058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.248234272003174,
+      "learning_rate": 4.549325931800159e-05,
+      "loss": 1.5147,
+      "step": 17059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8126657009124756,
+      "learning_rate": 4.549299497753106e-05,
+      "loss": 1.4844,
+      "step": 17060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6234310865402222,
+      "learning_rate": 4.5492730637060535e-05,
+      "loss": 1.3204,
+      "step": 17061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5669602155685425,
+      "learning_rate": 4.549246629659001e-05,
+      "loss": 1.3339,
+      "step": 17062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7099270820617676,
+      "learning_rate": 4.549220195611948e-05,
+      "loss": 1.5106,
+      "step": 17063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.545254230499268,
+      "learning_rate": 4.549193761564896e-05,
+      "loss": 1.3726,
+      "step": 17064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4340977668762207,
+      "learning_rate": 4.549167327517843e-05,
+      "loss": 1.2785,
+      "step": 17065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5894614458084106,
+      "learning_rate": 4.549140893470791e-05,
+      "loss": 1.3088,
+      "step": 17066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.855391025543213,
+      "learning_rate": 4.5491144594237376e-05,
+      "loss": 1.3786,
+      "step": 17067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6268445253372192,
+      "learning_rate": 4.5490880253766857e-05,
+      "loss": 1.3954,
+      "step": 17068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6159658432006836,
+      "learning_rate": 4.549061591329633e-05,
+      "loss": 1.3461,
+      "step": 17069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6813950538635254,
+      "learning_rate": 4.5490351572825804e-05,
+      "loss": 1.4174,
+      "step": 17070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7490812540054321,
+      "learning_rate": 4.549008723235528e-05,
+      "loss": 1.3624,
+      "step": 17071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6178967952728271,
+      "learning_rate": 4.548982289188475e-05,
+      "loss": 1.2705,
+      "step": 17072
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8203486204147339,
+      "learning_rate": 4.5489558551414224e-05,
+      "loss": 1.5786,
+      "step": 17073
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4404462575912476,
+      "learning_rate": 4.54892942109437e-05,
+      "loss": 2.1411,
+      "step": 17074
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.611716389656067,
+      "learning_rate": 4.548902987047317e-05,
+      "loss": 2.3245,
+      "step": 17075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3957902193069458,
+      "learning_rate": 4.5488765530002645e-05,
+      "loss": 1.4484,
+      "step": 17076
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0845229625701904,
+      "learning_rate": 4.548850118953212e-05,
+      "loss": 1.4354,
+      "step": 17077
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1800456047058105,
+      "learning_rate": 4.548823684906159e-05,
+      "loss": 1.5281,
+      "step": 17078
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9211275577545166,
+      "learning_rate": 4.548797250859107e-05,
+      "loss": 1.3841,
+      "step": 17079
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.253904104232788,
+      "learning_rate": 4.548770816812054e-05,
+      "loss": 1.536,
+      "step": 17080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6467422246932983,
+      "learning_rate": 4.548744382765002e-05,
+      "loss": 1.4187,
+      "step": 17081
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.081554651260376,
+      "learning_rate": 4.5487179487179487e-05,
+      "loss": 1.4777,
+      "step": 17082
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.830069899559021,
+      "learning_rate": 4.548691514670897e-05,
+      "loss": 1.4833,
+      "step": 17083
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5545605421066284,
+      "learning_rate": 4.548665080623844e-05,
+      "loss": 1.2287,
+      "step": 17084
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7617546319961548,
+      "learning_rate": 4.548638646576791e-05,
+      "loss": 1.4326,
+      "step": 17085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6609528064727783,
+      "learning_rate": 4.548612212529739e-05,
+      "loss": 1.2896,
+      "step": 17086
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.420180082321167,
+      "learning_rate": 4.5485857784826854e-05,
+      "loss": 1.4328,
+      "step": 17087
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6737608909606934,
+      "learning_rate": 4.5485593444356335e-05,
+      "loss": 1.4124,
+      "step": 17088
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.547763466835022,
+      "learning_rate": 4.548532910388581e-05,
+      "loss": 1.4034,
+      "step": 17089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8736048936843872,
+      "learning_rate": 4.548506476341528e-05,
+      "loss": 2.7497,
+      "step": 17090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.368827223777771,
+      "learning_rate": 4.5484800422944755e-05,
+      "loss": 1.1956,
+      "step": 17091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8648009300231934,
+      "learning_rate": 4.548453608247423e-05,
+      "loss": 1.5159,
+      "step": 17092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.976811408996582,
+      "learning_rate": 4.54842717420037e-05,
+      "loss": 1.3864,
+      "step": 17093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0240299701690674,
+      "learning_rate": 4.548400740153318e-05,
+      "loss": 1.2707,
+      "step": 17094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4897925853729248,
+      "learning_rate": 4.548374306106265e-05,
+      "loss": 1.993,
+      "step": 17095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9790321588516235,
+      "learning_rate": 4.548347872059212e-05,
+      "loss": 1.458,
+      "step": 17096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.757569670677185,
+      "learning_rate": 4.54832143801216e-05,
+      "loss": 1.4596,
+      "step": 17097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.537952184677124,
+      "learning_rate": 4.548295003965107e-05,
+      "loss": 1.638,
+      "step": 17098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4963572025299072,
+      "learning_rate": 4.548268569918055e-05,
+      "loss": 1.2165,
+      "step": 17099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.755547285079956,
+      "learning_rate": 4.548242135871002e-05,
+      "loss": 1.2221,
+      "step": 17100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5074148178100586,
+      "learning_rate": 4.54821570182395e-05,
+      "loss": 1.3653,
+      "step": 17101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0377197265625,
+      "learning_rate": 4.5481892677768965e-05,
+      "loss": 1.3835,
+      "step": 17102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.362159252166748,
+      "learning_rate": 4.5481628337298445e-05,
+      "loss": 1.3922,
+      "step": 17103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0942161083221436,
+      "learning_rate": 4.548136399682792e-05,
+      "loss": 1.3286,
+      "step": 17104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.765846848487854,
+      "learning_rate": 4.548109965635739e-05,
+      "loss": 1.4163,
+      "step": 17105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8655458688735962,
+      "learning_rate": 4.5480835315886866e-05,
+      "loss": 1.3488,
+      "step": 17106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0864126682281494,
+      "learning_rate": 4.548057097541634e-05,
+      "loss": 1.4843,
+      "step": 17107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.588568925857544,
+      "learning_rate": 4.548030663494581e-05,
+      "loss": 1.3881,
+      "step": 17108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0251426696777344,
+      "learning_rate": 4.5480042294475287e-05,
+      "loss": 1.2804,
+      "step": 17109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6748801469802856,
+      "learning_rate": 4.547977795400476e-05,
+      "loss": 1.3714,
+      "step": 17110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7470386028289795,
+      "learning_rate": 4.5479513613534234e-05,
+      "loss": 1.3776,
+      "step": 17111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7047048807144165,
+      "learning_rate": 4.547924927306371e-05,
+      "loss": 2.1204,
+      "step": 17112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8901938199996948,
+      "learning_rate": 4.547898493259318e-05,
+      "loss": 1.3689,
+      "step": 17113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9313946962356567,
+      "learning_rate": 4.547872059212266e-05,
+      "loss": 1.4761,
+      "step": 17114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.435123920440674,
+      "learning_rate": 4.547845625165213e-05,
+      "loss": 1.6484,
+      "step": 17115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0221145153045654,
+      "learning_rate": 4.547819191118161e-05,
+      "loss": 1.51,
+      "step": 17116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7739920616149902,
+      "learning_rate": 4.5477927570711075e-05,
+      "loss": 1.5104,
+      "step": 17117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9263972043991089,
+      "learning_rate": 4.5477663230240555e-05,
+      "loss": 1.4373,
+      "step": 17118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.748427391052246,
+      "learning_rate": 4.547739888977003e-05,
+      "loss": 1.3966,
+      "step": 17119
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6977038383483887,
+      "learning_rate": 4.54771345492995e-05,
+      "loss": 1.3202,
+      "step": 17120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8431117534637451,
+      "learning_rate": 4.5476870208828976e-05,
+      "loss": 1.3054,
+      "step": 17121
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5764573812484741,
+      "learning_rate": 4.547660586835844e-05,
+      "loss": 2.047,
+      "step": 17122
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8493173122406006,
+      "learning_rate": 4.547634152788792e-05,
+      "loss": 1.5797,
+      "step": 17123
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7855000495910645,
+      "learning_rate": 4.54760771874174e-05,
+      "loss": 1.4921,
+      "step": 17124
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.368865966796875,
+      "learning_rate": 4.547581284694687e-05,
+      "loss": 1.4974,
+      "step": 17125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7015266418457031,
+      "learning_rate": 4.5475548506476344e-05,
+      "loss": 1.271,
+      "step": 17126
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6094048023223877,
+      "learning_rate": 4.547528416600582e-05,
+      "loss": 1.4015,
+      "step": 17127
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4361683130264282,
+      "learning_rate": 4.547501982553529e-05,
+      "loss": 1.0368,
+      "step": 17128
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4289671182632446,
+      "learning_rate": 4.547475548506477e-05,
+      "loss": 1.2991,
+      "step": 17129
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2339088916778564,
+      "learning_rate": 4.547449114459424e-05,
+      "loss": 1.387,
+      "step": 17130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5554651021957397,
+      "learning_rate": 4.547422680412371e-05,
+      "loss": 1.4219,
+      "step": 17131
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1656036376953125,
+      "learning_rate": 4.5473962463653185e-05,
+      "loss": 1.4533,
+      "step": 17132
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2780349254608154,
+      "learning_rate": 4.547369812318266e-05,
+      "loss": 1.3764,
+      "step": 17133
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6828947067260742,
+      "learning_rate": 4.547343378271214e-05,
+      "loss": 1.3428,
+      "step": 17134
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.103121042251587,
+      "learning_rate": 4.5473169442241606e-05,
+      "loss": 1.3704,
+      "step": 17135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3897287845611572,
+      "learning_rate": 4.5472905101771087e-05,
+      "loss": 1.205,
+      "step": 17136
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6208770275115967,
+      "learning_rate": 4.547264076130055e-05,
+      "loss": 1.5995,
+      "step": 17137
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6407583951950073,
+      "learning_rate": 4.5472376420830034e-05,
+      "loss": 1.3012,
+      "step": 17138
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7270344495773315,
+      "learning_rate": 4.547211208035951e-05,
+      "loss": 1.402,
+      "step": 17139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4999618530273438,
+      "learning_rate": 4.547184773988898e-05,
+      "loss": 1.2056,
+      "step": 17140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.651293396949768,
+      "learning_rate": 4.5471583399418454e-05,
+      "loss": 1.4273,
+      "step": 17141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5811805725097656,
+      "learning_rate": 4.547131905894793e-05,
+      "loss": 1.3876,
+      "step": 17142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.916103720664978,
+      "learning_rate": 4.54710547184774e-05,
+      "loss": 1.4679,
+      "step": 17143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6669247150421143,
+      "learning_rate": 4.5470790378006875e-05,
+      "loss": 1.4732,
+      "step": 17144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.606658697128296,
+      "learning_rate": 4.547052603753635e-05,
+      "loss": 1.3603,
+      "step": 17145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2954185009002686,
+      "learning_rate": 4.547026169706582e-05,
+      "loss": 1.3271,
+      "step": 17146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9428695440292358,
+      "learning_rate": 4.5469997356595296e-05,
+      "loss": 1.5336,
+      "step": 17147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.355447292327881,
+      "learning_rate": 4.546973301612477e-05,
+      "loss": 1.3784,
+      "step": 17148
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.535679578781128,
+      "learning_rate": 4.546946867565425e-05,
+      "loss": 1.2627,
+      "step": 17149
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.46226966381073,
+      "learning_rate": 4.5469204335183716e-05,
+      "loss": 1.3525,
+      "step": 17150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6937049627304077,
+      "learning_rate": 4.54689399947132e-05,
+      "loss": 1.3175,
+      "step": 17151
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0334510803222656,
+      "learning_rate": 4.5468675654242664e-05,
+      "loss": 1.5085,
+      "step": 17152
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7767763137817383,
+      "learning_rate": 4.5468411313772144e-05,
+      "loss": 1.2464,
+      "step": 17153
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9919805526733398,
+      "learning_rate": 4.546814697330161e-05,
+      "loss": 1.3522,
+      "step": 17154
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8804460763931274,
+      "learning_rate": 4.546788263283109e-05,
+      "loss": 1.5593,
+      "step": 17155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5845855474472046,
+      "learning_rate": 4.5467618292360565e-05,
+      "loss": 1.2313,
+      "step": 17156
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6419775485992432,
+      "learning_rate": 4.546735395189004e-05,
+      "loss": 1.3358,
+      "step": 17157
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6574759483337402,
+      "learning_rate": 4.546708961141951e-05,
+      "loss": 1.4894,
+      "step": 17158
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8613307476043701,
+      "learning_rate": 4.546682527094898e-05,
+      "loss": 1.4838,
+      "step": 17159
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7542234659194946,
+      "learning_rate": 4.546656093047846e-05,
+      "loss": 1.2903,
+      "step": 17160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9820395708084106,
+      "learning_rate": 4.546629659000793e-05,
+      "loss": 1.3975,
+      "step": 17161
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5903702974319458,
+      "learning_rate": 4.5466032249537406e-05,
+      "loss": 2.3299,
+      "step": 17162
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.947054386138916,
+      "learning_rate": 4.546576790906688e-05,
+      "loss": 1.38,
+      "step": 17163
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.340538740158081,
+      "learning_rate": 4.546550356859635e-05,
+      "loss": 1.5212,
+      "step": 17164
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4347164630889893,
+      "learning_rate": 4.546523922812583e-05,
+      "loss": 1.435,
+      "step": 17165
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5890352725982666,
+      "learning_rate": 4.546497488765531e-05,
+      "loss": 1.4511,
+      "step": 17166
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3963663578033447,
+      "learning_rate": 4.5464710547184774e-05,
+      "loss": 1.6952,
+      "step": 17167
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.692384958267212,
+      "learning_rate": 4.546444620671425e-05,
+      "loss": 1.3364,
+      "step": 17168
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.535112142562866,
+      "learning_rate": 4.546418186624372e-05,
+      "loss": 1.5599,
+      "step": 17169
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.2727370262146,
+      "learning_rate": 4.5463917525773195e-05,
+      "loss": 1.4396,
+      "step": 17170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0709474086761475,
+      "learning_rate": 4.5463653185302675e-05,
+      "loss": 1.3318,
+      "step": 17171
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8289145231246948,
+      "learning_rate": 4.546338884483214e-05,
+      "loss": 1.377,
+      "step": 17172
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1470751762390137,
+      "learning_rate": 4.546312450436162e-05,
+      "loss": 1.4651,
+      "step": 17173
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7517515420913696,
+      "learning_rate": 4.546286016389109e-05,
+      "loss": 1.263,
+      "step": 17174
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5358259677886963,
+      "learning_rate": 4.546259582342057e-05,
+      "loss": 1.3179,
+      "step": 17175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6585488319396973,
+      "learning_rate": 4.546233148295004e-05,
+      "loss": 2.5806,
+      "step": 17176
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4730688333511353,
+      "learning_rate": 4.5462067142479516e-05,
+      "loss": 1.4027,
+      "step": 17177
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9223586320877075,
+      "learning_rate": 4.546180280200899e-05,
+      "loss": 1.4151,
+      "step": 17178
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9246872663497925,
+      "learning_rate": 4.5461538461538464e-05,
+      "loss": 1.3568,
+      "step": 17179
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.826451063156128,
+      "learning_rate": 4.546127412106794e-05,
+      "loss": 1.3402,
+      "step": 17180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6608421802520752,
+      "learning_rate": 4.546100978059741e-05,
+      "loss": 1.3515,
+      "step": 17181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8701733350753784,
+      "learning_rate": 4.5460745440126884e-05,
+      "loss": 1.2529,
+      "step": 17182
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7267577648162842,
+      "learning_rate": 4.546048109965636e-05,
+      "loss": 1.3735,
+      "step": 17183
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8426353931427002,
+      "learning_rate": 4.546021675918583e-05,
+      "loss": 1.3299,
+      "step": 17184
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3080925941467285,
+      "learning_rate": 4.5459952418715305e-05,
+      "loss": 1.3772,
+      "step": 17185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7997804880142212,
+      "learning_rate": 4.5459688078244785e-05,
+      "loss": 1.3828,
+      "step": 17186
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7334270477294922,
+      "learning_rate": 4.545942373777425e-05,
+      "loss": 1.4598,
+      "step": 17187
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9885172843933105,
+      "learning_rate": 4.545915939730373e-05,
+      "loss": 1.5718,
+      "step": 17188
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3320624828338623,
+      "learning_rate": 4.54588950568332e-05,
+      "loss": 1.2449,
+      "step": 17189
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5921244621276855,
+      "learning_rate": 4.545863071636268e-05,
+      "loss": 1.2938,
+      "step": 17190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7395179271697998,
+      "learning_rate": 4.545836637589215e-05,
+      "loss": 1.3062,
+      "step": 17191
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9117897748947144,
+      "learning_rate": 4.545810203542163e-05,
+      "loss": 1.4489,
+      "step": 17192
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5152369737625122,
+      "learning_rate": 4.54578376949511e-05,
+      "loss": 1.2259,
+      "step": 17193
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5521669387817383,
+      "learning_rate": 4.545757335448057e-05,
+      "loss": 1.6172,
+      "step": 17194
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5332698822021484,
+      "learning_rate": 4.545730901401005e-05,
+      "loss": 1.508,
+      "step": 17195
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6697678565979004,
+      "learning_rate": 4.545704467353952e-05,
+      "loss": 1.4263,
+      "step": 17196
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.058115005493164,
+      "learning_rate": 4.5456780333068995e-05,
+      "loss": 1.3989,
+      "step": 17197
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9008485078811646,
+      "learning_rate": 4.545651599259847e-05,
+      "loss": 1.2315,
+      "step": 17198
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9955469369888306,
+      "learning_rate": 4.545625165212794e-05,
+      "loss": 1.3361,
+      "step": 17199
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7497594356536865,
+      "learning_rate": 4.5455987311657415e-05,
+      "loss": 1.3245,
+      "step": 17200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.382163405418396,
+      "learning_rate": 4.5455722971186896e-05,
+      "loss": 1.2806,
+      "step": 17201
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5297683477401733,
+      "learning_rate": 4.545545863071636e-05,
+      "loss": 1.3122,
+      "step": 17202
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.686246633529663,
+      "learning_rate": 4.5455194290245836e-05,
+      "loss": 1.4585,
+      "step": 17203
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6976208686828613,
+      "learning_rate": 4.545492994977531e-05,
+      "loss": 1.34,
+      "step": 17204
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8604276180267334,
+      "learning_rate": 4.545466560930478e-05,
+      "loss": 1.3157,
+      "step": 17205
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.674731969833374,
+      "learning_rate": 4.5454401268834264e-05,
+      "loss": 1.3616,
+      "step": 17206
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.04191255569458,
+      "learning_rate": 4.545413692836373e-05,
+      "loss": 2.0176,
+      "step": 17207
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.812343120574951,
+      "learning_rate": 4.545387258789321e-05,
+      "loss": 2.8594,
+      "step": 17208
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.892565131187439,
+      "learning_rate": 4.545360824742268e-05,
+      "loss": 1.5746,
+      "step": 17209
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.166125774383545,
+      "learning_rate": 4.545334390695216e-05,
+      "loss": 1.4698,
+      "step": 17210
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0895724296569824,
+      "learning_rate": 4.545307956648163e-05,
+      "loss": 1.4596,
+      "step": 17211
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.306910276412964,
+      "learning_rate": 4.5452815226011105e-05,
+      "loss": 1.4574,
+      "step": 17212
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8559576272964478,
+      "learning_rate": 4.545255088554058e-05,
+      "loss": 1.4059,
+      "step": 17213
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8120776414871216,
+      "learning_rate": 4.545228654507005e-05,
+      "loss": 1.2897,
+      "step": 17214
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8544824123382568,
+      "learning_rate": 4.5452022204599526e-05,
+      "loss": 1.3684,
+      "step": 17215
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6465423107147217,
+      "learning_rate": 4.5451757864129e-05,
+      "loss": 1.2064,
+      "step": 17216
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7621253728866577,
+      "learning_rate": 4.545149352365847e-05,
+      "loss": 1.3763,
+      "step": 17217
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5720300674438477,
+      "learning_rate": 4.5451229183187946e-05,
+      "loss": 1.3791,
+      "step": 17218
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9784739017486572,
+      "learning_rate": 4.545096484271742e-05,
+      "loss": 1.548,
+      "step": 17219
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8799073696136475,
+      "learning_rate": 4.5450700502246894e-05,
+      "loss": 1.4204,
+      "step": 17220
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.456634521484375,
+      "learning_rate": 4.5450436161776374e-05,
+      "loss": 1.3666,
+      "step": 17221
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.45549738407135,
+      "learning_rate": 4.545017182130584e-05,
+      "loss": 1.4617,
+      "step": 17222
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.001168966293335,
+      "learning_rate": 4.544990748083532e-05,
+      "loss": 1.3009,
+      "step": 17223
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.694467544555664,
+      "learning_rate": 4.544964314036479e-05,
+      "loss": 1.5463,
+      "step": 17224
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.108689785003662,
+      "learning_rate": 4.544937879989427e-05,
+      "loss": 1.5267,
+      "step": 17225
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.001009225845337,
+      "learning_rate": 4.544911445942374e-05,
+      "loss": 2.1619,
+      "step": 17226
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6685389280319214,
+      "learning_rate": 4.5448850118953215e-05,
+      "loss": 1.2803,
+      "step": 17227
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9557864665985107,
+      "learning_rate": 4.544858577848269e-05,
+      "loss": 1.2857,
+      "step": 17228
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.67299485206604,
+      "learning_rate": 4.544832143801216e-05,
+      "loss": 1.5112,
+      "step": 17229
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7432469129562378,
+      "learning_rate": 4.5448057097541636e-05,
+      "loss": 1.3299,
+      "step": 17230
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7641443014144897,
+      "learning_rate": 4.544779275707111e-05,
+      "loss": 1.4344,
+      "step": 17231
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.061997890472412,
+      "learning_rate": 4.544752841660058e-05,
+      "loss": 1.3954,
+      "step": 17232
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.793213129043579,
+      "learning_rate": 4.544726407613006e-05,
+      "loss": 1.4299,
+      "step": 17233
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.129513740539551,
+      "learning_rate": 4.544699973565953e-05,
+      "loss": 1.6366,
+      "step": 17234
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6830322742462158,
+      "learning_rate": 4.5446735395189004e-05,
+      "loss": 1.3766,
+      "step": 17235
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6403008699417114,
+      "learning_rate": 4.5446471054718484e-05,
+      "loss": 1.2851,
+      "step": 17236
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4921294450759888,
+      "learning_rate": 4.544620671424795e-05,
+      "loss": 1.28,
+      "step": 17237
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.743226170539856,
+      "learning_rate": 4.544594237377743e-05,
+      "loss": 1.4234,
+      "step": 17238
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5833923816680908,
+      "learning_rate": 4.54456780333069e-05,
+      "loss": 1.3443,
+      "step": 17239
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.751549482345581,
+      "learning_rate": 4.544541369283637e-05,
+      "loss": 1.4118,
+      "step": 17240
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9046939611434937,
+      "learning_rate": 4.544514935236585e-05,
+      "loss": 1.6714,
+      "step": 17241
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6928882598876953,
+      "learning_rate": 4.544488501189532e-05,
+      "loss": 1.5456,
+      "step": 17242
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6516215801239014,
+      "learning_rate": 4.54446206714248e-05,
+      "loss": 1.3732,
+      "step": 17243
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7901246547698975,
+      "learning_rate": 4.5444356330954266e-05,
+      "loss": 1.4185,
+      "step": 17244
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5857340097427368,
+      "learning_rate": 4.5444091990483746e-05,
+      "loss": 1.3313,
+      "step": 17245
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8460637331008911,
+      "learning_rate": 4.544382765001322e-05,
+      "loss": 1.1852,
+      "step": 17246
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7727316617965698,
+      "learning_rate": 4.5443563309542694e-05,
+      "loss": 1.3108,
+      "step": 17247
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9913325309753418,
+      "learning_rate": 4.544329896907217e-05,
+      "loss": 1.3739,
+      "step": 17248
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7366918325424194,
+      "learning_rate": 4.544303462860164e-05,
+      "loss": 1.3262,
+      "step": 17249
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7387555837631226,
+      "learning_rate": 4.5442770288131114e-05,
+      "loss": 1.425,
+      "step": 17250
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6875871419906616,
+      "learning_rate": 4.544250594766059e-05,
+      "loss": 1.3832,
+      "step": 17251
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4960517883300781,
+      "learning_rate": 4.544224160719006e-05,
+      "loss": 1.1546,
+      "step": 17252
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.625180959701538,
+      "learning_rate": 4.5441977266719535e-05,
+      "loss": 1.2744,
+      "step": 17253
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9591224193572998,
+      "learning_rate": 4.544171292624901e-05,
+      "loss": 1.4349,
+      "step": 17254
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7417716979980469,
+      "learning_rate": 4.544144858577848e-05,
+      "loss": 1.3913,
+      "step": 17255
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.888861060142517,
+      "learning_rate": 4.544118424530796e-05,
+      "loss": 1.276,
+      "step": 17256
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9746679067611694,
+      "learning_rate": 4.544091990483743e-05,
+      "loss": 1.4886,
+      "step": 17257
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9904241561889648,
+      "learning_rate": 4.544065556436691e-05,
+      "loss": 1.4961,
+      "step": 17258
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6927118301391602,
+      "learning_rate": 4.5440391223896376e-05,
+      "loss": 1.5929,
+      "step": 17259
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.275024652481079,
+      "learning_rate": 4.544012688342586e-05,
+      "loss": 1.5039,
+      "step": 17260
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.46835994720459,
+      "learning_rate": 4.543986254295533e-05,
+      "loss": 1.6088,
+      "step": 17261
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8845235109329224,
+      "learning_rate": 4.5439598202484804e-05,
+      "loss": 1.49,
+      "step": 17262
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.507028579711914,
+      "learning_rate": 4.543933386201428e-05,
+      "loss": 1.3977,
+      "step": 17263
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9625126123428345,
+      "learning_rate": 4.543906952154375e-05,
+      "loss": 1.4912,
+      "step": 17264
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.484046220779419,
+      "learning_rate": 4.5438805181073225e-05,
+      "loss": 1.3816,
+      "step": 17265
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6133440732955933,
+      "learning_rate": 4.54385408406027e-05,
+      "loss": 1.3868,
+      "step": 17266
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9640216827392578,
+      "learning_rate": 4.543827650013217e-05,
+      "loss": 1.7607,
+      "step": 17267
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.734523296356201,
+      "learning_rate": 4.5438012159661645e-05,
+      "loss": 1.576,
+      "step": 17268
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8163286447525024,
+      "learning_rate": 4.543774781919112e-05,
+      "loss": 1.4078,
+      "step": 17269
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.859001874923706,
+      "learning_rate": 4.543748347872059e-05,
+      "loss": 1.3524,
+      "step": 17270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3242452144622803,
+      "learning_rate": 4.543721913825007e-05,
+      "loss": 1.4415,
+      "step": 17271
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7698757648468018,
+      "learning_rate": 4.543695479777954e-05,
+      "loss": 1.3516,
+      "step": 17272
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.324733853340149,
+      "learning_rate": 4.543669045730902e-05,
+      "loss": 1.207,
+      "step": 17273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5310288667678833,
+      "learning_rate": 4.543642611683849e-05,
+      "loss": 1.4209,
+      "step": 17274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9828004837036133,
+      "learning_rate": 4.543616177636796e-05,
+      "loss": 1.4412,
+      "step": 17275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.219653367996216,
+      "learning_rate": 4.543589743589744e-05,
+      "loss": 1.527,
+      "step": 17276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3242950439453125,
+      "learning_rate": 4.543563309542691e-05,
+      "loss": 1.1588,
+      "step": 17277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8274422883987427,
+      "learning_rate": 4.543536875495639e-05,
+      "loss": 1.3965,
+      "step": 17278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3269999027252197,
+      "learning_rate": 4.5435104414485855e-05,
+      "loss": 1.2337,
+      "step": 17279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7947825193405151,
+      "learning_rate": 4.5434840074015335e-05,
+      "loss": 1.4354,
+      "step": 17280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1872599124908447,
+      "learning_rate": 4.543457573354481e-05,
+      "loss": 1.6033,
+      "step": 17281
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.353062868118286,
+      "learning_rate": 4.543431139307428e-05,
+      "loss": 1.5149,
+      "step": 17282
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9715871810913086,
+      "learning_rate": 4.5434047052603756e-05,
+      "loss": 1.4958,
+      "step": 17283
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.961365818977356,
+      "learning_rate": 4.543378271213323e-05,
+      "loss": 1.4196,
+      "step": 17284
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4560998678207397,
+      "learning_rate": 4.54335183716627e-05,
+      "loss": 1.4137,
+      "step": 17285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0578060150146484,
+      "learning_rate": 4.5433254031192176e-05,
+      "loss": 1.349,
+      "step": 17286
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0958590507507324,
+      "learning_rate": 4.543298969072165e-05,
+      "loss": 1.467,
+      "step": 17287
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.616757869720459,
+      "learning_rate": 4.5432725350251123e-05,
+      "loss": 1.3583,
+      "step": 17288
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5917575359344482,
+      "learning_rate": 4.54324610097806e-05,
+      "loss": 1.3194,
+      "step": 17289
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0194199085235596,
+      "learning_rate": 4.543219666931007e-05,
+      "loss": 1.4949,
+      "step": 17290
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7540620565414429,
+      "learning_rate": 4.543193232883955e-05,
+      "loss": 1.3243,
+      "step": 17291
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1549313068389893,
+      "learning_rate": 4.543166798836902e-05,
+      "loss": 1.3606,
+      "step": 17292
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0141854286193848,
+      "learning_rate": 4.54314036478985e-05,
+      "loss": 1.4878,
+      "step": 17293
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7344169616699219,
+      "learning_rate": 4.5431139307427965e-05,
+      "loss": 1.3699,
+      "step": 17294
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8099061250686646,
+      "learning_rate": 4.5430874966957445e-05,
+      "loss": 1.3062,
+      "step": 17295
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.978509783744812,
+      "learning_rate": 4.543061062648692e-05,
+      "loss": 1.2693,
+      "step": 17296
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6047818660736084,
+      "learning_rate": 4.543034628601639e-05,
+      "loss": 1.2914,
+      "step": 17297
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6210821866989136,
+      "learning_rate": 4.5430081945545866e-05,
+      "loss": 1.4396,
+      "step": 17298
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4377280473709106,
+      "learning_rate": 4.542981760507534e-05,
+      "loss": 1.3605,
+      "step": 17299
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7368264198303223,
+      "learning_rate": 4.542955326460481e-05,
+      "loss": 1.2924,
+      "step": 17300
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4962172508239746,
+      "learning_rate": 4.542928892413429e-05,
+      "loss": 1.6057,
+      "step": 17301
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5632805824279785,
+      "learning_rate": 4.542902458366376e-05,
+      "loss": 1.5327,
+      "step": 17302
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6490364074707031,
+      "learning_rate": 4.5428760243193234e-05,
+      "loss": 1.3011,
+      "step": 17303
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2202694416046143,
+      "learning_rate": 4.542849590272271e-05,
+      "loss": 1.4273,
+      "step": 17304
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2152185440063477,
+      "learning_rate": 4.542823156225218e-05,
+      "loss": 1.4768,
+      "step": 17305
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7113832235336304,
+      "learning_rate": 4.542796722178166e-05,
+      "loss": 1.4118,
+      "step": 17306
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.974632501602173,
+      "learning_rate": 4.542770288131113e-05,
+      "loss": 1.5228,
+      "step": 17307
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7861289978027344,
+      "learning_rate": 4.542743854084061e-05,
+      "loss": 1.3487,
+      "step": 17308
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8856500387191772,
+      "learning_rate": 4.5427174200370075e-05,
+      "loss": 1.4241,
+      "step": 17309
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2947278022766113,
+      "learning_rate": 4.5426909859899556e-05,
+      "loss": 2.2868,
+      "step": 17310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5455503463745117,
+      "learning_rate": 4.542664551942903e-05,
+      "loss": 1.3329,
+      "step": 17311
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0278546810150146,
+      "learning_rate": 4.5426381178958496e-05,
+      "loss": 1.505,
+      "step": 17312
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6203421354293823,
+      "learning_rate": 4.5426116838487976e-05,
+      "loss": 1.3587,
+      "step": 17313
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.661606788635254,
+      "learning_rate": 4.542585249801744e-05,
+      "loss": 1.2871,
+      "step": 17314
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8144031763076782,
+      "learning_rate": 4.5425588157546923e-05,
+      "loss": 1.4014,
+      "step": 17315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9019906520843506,
+      "learning_rate": 4.54253238170764e-05,
+      "loss": 1.4879,
+      "step": 17316
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.196786880493164,
+      "learning_rate": 4.542505947660587e-05,
+      "loss": 1.614,
+      "step": 17317
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6151355504989624,
+      "learning_rate": 4.5424795136135344e-05,
+      "loss": 1.4462,
+      "step": 17318
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.685798168182373,
+      "learning_rate": 4.542453079566482e-05,
+      "loss": 1.4378,
+      "step": 17319
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6865901947021484,
+      "learning_rate": 4.542426645519429e-05,
+      "loss": 1.2248,
+      "step": 17320
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9496029615402222,
+      "learning_rate": 4.5424002114723765e-05,
+      "loss": 1.5596,
+      "step": 17321
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5095648765563965,
+      "learning_rate": 4.542373777425324e-05,
+      "loss": 1.5438,
+      "step": 17322
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.217970132827759,
+      "learning_rate": 4.542347343378271e-05,
+      "loss": 1.4217,
+      "step": 17323
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.748306155204773,
+      "learning_rate": 4.5423209093312186e-05,
+      "loss": 1.7812,
+      "step": 17324
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0353450775146484,
+      "learning_rate": 4.542294475284166e-05,
+      "loss": 1.492,
+      "step": 17325
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.976271390914917,
+      "learning_rate": 4.542268041237114e-05,
+      "loss": 1.4715,
+      "step": 17326
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5331552028656006,
+      "learning_rate": 4.5422416071900606e-05,
+      "loss": 1.3452,
+      "step": 17327
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7048360109329224,
+      "learning_rate": 4.542215173143009e-05,
+      "loss": 1.512,
+      "step": 17328
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.966370701789856,
+      "learning_rate": 4.5421887390959553e-05,
+      "loss": 1.3877,
+      "step": 17329
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8780014514923096,
+      "learning_rate": 4.5421623050489034e-05,
+      "loss": 1.5299,
+      "step": 17330
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8231585025787354,
+      "learning_rate": 4.542135871001851e-05,
+      "loss": 2.4423,
+      "step": 17331
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8940552473068237,
+      "learning_rate": 4.542109436954798e-05,
+      "loss": 1.5245,
+      "step": 17332
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5556572675704956,
+      "learning_rate": 4.5420830029077455e-05,
+      "loss": 1.3457,
+      "step": 17333
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4831116199493408,
+      "learning_rate": 4.542056568860693e-05,
+      "loss": 1.3682,
+      "step": 17334
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0868570804595947,
+      "learning_rate": 4.54203013481364e-05,
+      "loss": 1.4601,
+      "step": 17335
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.80068838596344,
+      "learning_rate": 4.5420037007665875e-05,
+      "loss": 1.3919,
+      "step": 17336
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.013089656829834,
+      "learning_rate": 4.541977266719535e-05,
+      "loss": 1.3286,
+      "step": 17337
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.200434446334839,
+      "learning_rate": 4.541950832672482e-05,
+      "loss": 1.4263,
+      "step": 17338
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0171847343444824,
+      "learning_rate": 4.5419243986254296e-05,
+      "loss": 1.4384,
+      "step": 17339
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7627482414245605,
+      "learning_rate": 4.541897964578377e-05,
+      "loss": 1.8508,
+      "step": 17340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8184233903884888,
+      "learning_rate": 4.541871530531325e-05,
+      "loss": 1.2459,
+      "step": 17341
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6030553579330444,
+      "learning_rate": 4.541845096484272e-05,
+      "loss": 1.2902,
+      "step": 17342
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6423667669296265,
+      "learning_rate": 4.54181866243722e-05,
+      "loss": 1.3149,
+      "step": 17343
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.96512770652771,
+      "learning_rate": 4.5417922283901664e-05,
+      "loss": 1.3332,
+      "step": 17344
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8493372201919556,
+      "learning_rate": 4.5417657943431144e-05,
+      "loss": 1.5467,
+      "step": 17345
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8232284784317017,
+      "learning_rate": 4.541739360296062e-05,
+      "loss": 1.4105,
+      "step": 17346
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.124135971069336,
+      "learning_rate": 4.541712926249009e-05,
+      "loss": 1.3058,
+      "step": 17347
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8672388792037964,
+      "learning_rate": 4.5416864922019565e-05,
+      "loss": 1.4294,
+      "step": 17348
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4381908178329468,
+      "learning_rate": 4.541660058154903e-05,
+      "loss": 1.3726,
+      "step": 17349
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4936208724975586,
+      "learning_rate": 4.541633624107851e-05,
+      "loss": 1.7619,
+      "step": 17350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0515079498291016,
+      "learning_rate": 4.5416071900607986e-05,
+      "loss": 1.3464,
+      "step": 17351
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.652250051498413,
+      "learning_rate": 4.541580756013746e-05,
+      "loss": 1.3633,
+      "step": 17352
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0606935024261475,
+      "learning_rate": 4.541554321966693e-05,
+      "loss": 1.3368,
+      "step": 17353
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5609636306762695,
+      "learning_rate": 4.5415278879196406e-05,
+      "loss": 1.4768,
+      "step": 17354
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4756513833999634,
+      "learning_rate": 4.541501453872588e-05,
+      "loss": 1.4194,
+      "step": 17355
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5415171384811401,
+      "learning_rate": 4.541475019825536e-05,
+      "loss": 1.2533,
+      "step": 17356
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.753286123275757,
+      "learning_rate": 4.541448585778483e-05,
+      "loss": 1.5299,
+      "step": 17357
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.882064938545227,
+      "learning_rate": 4.54142215173143e-05,
+      "loss": 1.3151,
+      "step": 17358
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.890774965286255,
+      "learning_rate": 4.5413957176843774e-05,
+      "loss": 1.5664,
+      "step": 17359
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.012204647064209,
+      "learning_rate": 4.541369283637325e-05,
+      "loss": 1.4298,
+      "step": 17360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6376217603683472,
+      "learning_rate": 4.541342849590273e-05,
+      "loss": 1.4031,
+      "step": 17361
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3725430965423584,
+      "learning_rate": 4.5413164155432195e-05,
+      "loss": 1.3857,
+      "step": 17362
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7345023155212402,
+      "learning_rate": 4.5412899814961675e-05,
+      "loss": 1.3682,
+      "step": 17363
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4284864664077759,
+      "learning_rate": 4.541263547449114e-05,
+      "loss": 1.3494,
+      "step": 17364
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6925748586654663,
+      "learning_rate": 4.541237113402062e-05,
+      "loss": 1.227,
+      "step": 17365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5891306400299072,
+      "learning_rate": 4.5412106793550096e-05,
+      "loss": 1.5453,
+      "step": 17366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6963599920272827,
+      "learning_rate": 4.541184245307957e-05,
+      "loss": 1.332,
+      "step": 17367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8014509677886963,
+      "learning_rate": 4.541157811260904e-05,
+      "loss": 1.3567,
+      "step": 17368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9686782360076904,
+      "learning_rate": 4.541131377213852e-05,
+      "loss": 1.3227,
+      "step": 17369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.817301630973816,
+      "learning_rate": 4.541104943166799e-05,
+      "loss": 1.4465,
+      "step": 17370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3450496196746826,
+      "learning_rate": 4.5410785091197464e-05,
+      "loss": 1.3944,
+      "step": 17371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8244584798812866,
+      "learning_rate": 4.541052075072694e-05,
+      "loss": 1.3204,
+      "step": 17372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8300424814224243,
+      "learning_rate": 4.541025641025641e-05,
+      "loss": 1.3814,
+      "step": 17373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.792452335357666,
+      "learning_rate": 4.5409992069785884e-05,
+      "loss": 1.459,
+      "step": 17374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8584192991256714,
+      "learning_rate": 4.540972772931536e-05,
+      "loss": 1.4975,
+      "step": 17375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6331299543380737,
+      "learning_rate": 4.540946338884484e-05,
+      "loss": 1.161,
+      "step": 17376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8576475381851196,
+      "learning_rate": 4.5409199048374305e-05,
+      "loss": 1.6043,
+      "step": 17377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.200359582901001,
+      "learning_rate": 4.5408934707903786e-05,
+      "loss": 1.3357,
+      "step": 17378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5173065662384033,
+      "learning_rate": 4.540867036743325e-05,
+      "loss": 2.174,
+      "step": 17379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6321672201156616,
+      "learning_rate": 4.540840602696273e-05,
+      "loss": 1.5394,
+      "step": 17380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1125946044921875,
+      "learning_rate": 4.5408141686492206e-05,
+      "loss": 1.4055,
+      "step": 17381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.011242628097534,
+      "learning_rate": 4.540787734602168e-05,
+      "loss": 1.4146,
+      "step": 17382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4595805406570435,
+      "learning_rate": 4.540761300555115e-05,
+      "loss": 1.2411,
+      "step": 17383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5190492868423462,
+      "learning_rate": 4.540734866508062e-05,
+      "loss": 1.264,
+      "step": 17384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5262610912322998,
+      "learning_rate": 4.54070843246101e-05,
+      "loss": 1.3337,
+      "step": 17385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6974501609802246,
+      "learning_rate": 4.5406819984139574e-05,
+      "loss": 1.4683,
+      "step": 17386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9219138622283936,
+      "learning_rate": 4.540655564366905e-05,
+      "loss": 1.3301,
+      "step": 17387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8857108354568481,
+      "learning_rate": 4.540629130319852e-05,
+      "loss": 1.4113,
+      "step": 17388
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.016529083251953,
+      "learning_rate": 4.5406026962727995e-05,
+      "loss": 1.3594,
+      "step": 17389
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1605045795440674,
+      "learning_rate": 4.540576262225747e-05,
+      "loss": 1.3234,
+      "step": 17390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4958254098892212,
+      "learning_rate": 4.540549828178695e-05,
+      "loss": 1.1798,
+      "step": 17391
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1363019943237305,
+      "learning_rate": 4.5405233941316416e-05,
+      "loss": 1.3364,
+      "step": 17392
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5195469856262207,
+      "learning_rate": 4.540496960084589e-05,
+      "loss": 1.4346,
+      "step": 17393
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0965847969055176,
+      "learning_rate": 4.540470526037536e-05,
+      "loss": 1.5324,
+      "step": 17394
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7142199277877808,
+      "learning_rate": 4.5404440919904836e-05,
+      "loss": 1.4516,
+      "step": 17395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4796799421310425,
+      "learning_rate": 4.5404176579434317e-05,
+      "loss": 1.3687,
+      "step": 17396
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5332823991775513,
+      "learning_rate": 4.540391223896378e-05,
+      "loss": 1.3629,
+      "step": 17397
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8398993015289307,
+      "learning_rate": 4.5403647898493264e-05,
+      "loss": 1.8253,
+      "step": 17398
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7148672342300415,
+      "learning_rate": 4.540338355802273e-05,
+      "loss": 1.1949,
+      "step": 17399
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6396899223327637,
+      "learning_rate": 4.540311921755221e-05,
+      "loss": 1.6142,
+      "step": 17400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7679494619369507,
+      "learning_rate": 4.5402854877081684e-05,
+      "loss": 1.3685,
+      "step": 17401
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.606844425201416,
+      "learning_rate": 4.540259053661116e-05,
+      "loss": 1.2702,
+      "step": 17402
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.604290246963501,
+      "learning_rate": 4.540232619614063e-05,
+      "loss": 1.5799,
+      "step": 17403
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.694047451019287,
+      "learning_rate": 4.5402061855670105e-05,
+      "loss": 1.495,
+      "step": 17404
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9015226364135742,
+      "learning_rate": 4.540179751519958e-05,
+      "loss": 1.5071,
+      "step": 17405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6242643594741821,
+      "learning_rate": 4.540153317472905e-05,
+      "loss": 1.3049,
+      "step": 17406
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6343330144882202,
+      "learning_rate": 4.5401268834258526e-05,
+      "loss": 1.3222,
+      "step": 17407
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5095323324203491,
+      "learning_rate": 4.5401004493788e-05,
+      "loss": 1.4224,
+      "step": 17408
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1648614406585693,
+      "learning_rate": 4.540074015331747e-05,
+      "loss": 2.5141,
+      "step": 17409
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.175466775894165,
+      "learning_rate": 4.5400475812846947e-05,
+      "loss": 1.4584,
+      "step": 17410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3964338302612305,
+      "learning_rate": 4.540021147237643e-05,
+      "loss": 1.5124,
+      "step": 17411
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5631169080734253,
+      "learning_rate": 4.5399947131905894e-05,
+      "loss": 1.3215,
+      "step": 17412
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7858986854553223,
+      "learning_rate": 4.5399682791435374e-05,
+      "loss": 1.4522,
+      "step": 17413
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0502748489379883,
+      "learning_rate": 4.539941845096484e-05,
+      "loss": 1.5282,
+      "step": 17414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.930076241493225,
+      "learning_rate": 4.539915411049432e-05,
+      "loss": 1.3777,
+      "step": 17415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0124032497406006,
+      "learning_rate": 4.5398889770023795e-05,
+      "loss": 1.3594,
+      "step": 17416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0469956398010254,
+      "learning_rate": 4.539862542955327e-05,
+      "loss": 1.2845,
+      "step": 17417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7021883726119995,
+      "learning_rate": 4.539836108908274e-05,
+      "loss": 2.0442,
+      "step": 17418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4930843114852905,
+      "learning_rate": 4.5398096748612215e-05,
+      "loss": 1.625,
+      "step": 17419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.769528865814209,
+      "learning_rate": 4.539783240814169e-05,
+      "loss": 1.2454,
+      "step": 17420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4140830039978027,
+      "learning_rate": 4.539756806767116e-05,
+      "loss": 1.4911,
+      "step": 17421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6145681142807007,
+      "learning_rate": 4.5397303727200636e-05,
+      "loss": 1.4163,
+      "step": 17422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0328550338745117,
+      "learning_rate": 4.539703938673011e-05,
+      "loss": 1.4377,
+      "step": 17423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0159120559692383,
+      "learning_rate": 4.539677504625958e-05,
+      "loss": 1.4229,
+      "step": 17424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6340882778167725,
+      "learning_rate": 4.539651070578906e-05,
+      "loss": 1.4078,
+      "step": 17425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8322646617889404,
+      "learning_rate": 4.539624636531854e-05,
+      "loss": 1.5279,
+      "step": 17426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4750897884368896,
+      "learning_rate": 4.5395982024848004e-05,
+      "loss": 1.2603,
+      "step": 17427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.795044183731079,
+      "learning_rate": 4.5395717684377484e-05,
+      "loss": 1.4677,
+      "step": 17428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8481550216674805,
+      "learning_rate": 4.539545334390695e-05,
+      "loss": 1.3533,
+      "step": 17429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5874775648117065,
+      "learning_rate": 4.5395189003436425e-05,
+      "loss": 1.2778,
+      "step": 17430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7862200736999512,
+      "learning_rate": 4.5394924662965905e-05,
+      "loss": 1.3748,
+      "step": 17431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.51091468334198,
+      "learning_rate": 4.539466032249537e-05,
+      "loss": 1.3982,
+      "step": 17432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.573901653289795,
+      "learning_rate": 4.539439598202485e-05,
+      "loss": 1.3561,
+      "step": 17433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6819190979003906,
+      "learning_rate": 4.539413164155432e-05,
+      "loss": 1.37,
+      "step": 17434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.330960512161255,
+      "learning_rate": 4.53938673010838e-05,
+      "loss": 1.5611,
+      "step": 17435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7871712446212769,
+      "learning_rate": 4.539360296061327e-05,
+      "loss": 1.4562,
+      "step": 17436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9886165857315063,
+      "learning_rate": 4.5393338620142747e-05,
+      "loss": 1.5363,
+      "step": 17437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.607118844985962,
+      "learning_rate": 4.539307427967222e-05,
+      "loss": 1.4884,
+      "step": 17438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.738069772720337,
+      "learning_rate": 4.5392809939201694e-05,
+      "loss": 1.3766,
+      "step": 17439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2215416431427,
+      "learning_rate": 4.539254559873117e-05,
+      "loss": 1.5414,
+      "step": 17440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5062980651855469,
+      "learning_rate": 4.539228125826064e-05,
+      "loss": 1.2608,
+      "step": 17441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.639025092124939,
+      "learning_rate": 4.5392016917790114e-05,
+      "loss": 1.8787,
+      "step": 17442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5303055047988892,
+      "learning_rate": 4.539175257731959e-05,
+      "loss": 1.4547,
+      "step": 17443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5895681381225586,
+      "learning_rate": 4.539148823684906e-05,
+      "loss": 1.2348,
+      "step": 17444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1314537525177,
+      "learning_rate": 4.5391223896378535e-05,
+      "loss": 1.4726,
+      "step": 17445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6590529680252075,
+      "learning_rate": 4.5390959555908015e-05,
+      "loss": 1.3904,
+      "step": 17446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4178544282913208,
+      "learning_rate": 4.539069521543748e-05,
+      "loss": 1.2703,
+      "step": 17447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7616181373596191,
+      "learning_rate": 4.539043087496696e-05,
+      "loss": 1.5018,
+      "step": 17448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8975824117660522,
+      "learning_rate": 4.539016653449643e-05,
+      "loss": 1.4186,
+      "step": 17449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.273096799850464,
+      "learning_rate": 4.538990219402591e-05,
+      "loss": 1.5802,
+      "step": 17450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8972065448760986,
+      "learning_rate": 4.538963785355538e-05,
+      "loss": 1.5405,
+      "step": 17451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2634046077728271,
+      "learning_rate": 4.538937351308486e-05,
+      "loss": 1.2933,
+      "step": 17452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.452006220817566,
+      "learning_rate": 4.538910917261433e-05,
+      "loss": 1.2287,
+      "step": 17453
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4866787195205688,
+      "learning_rate": 4.5388844832143804e-05,
+      "loss": 1.1452,
+      "step": 17454
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9801884889602661,
+      "learning_rate": 4.538858049167328e-05,
+      "loss": 1.5261,
+      "step": 17455
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.630937099456787,
+      "learning_rate": 4.538831615120275e-05,
+      "loss": 1.3954,
+      "step": 17456
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6157666444778442,
+      "learning_rate": 4.5388051810732225e-05,
+      "loss": 1.0944,
+      "step": 17457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9934507608413696,
+      "learning_rate": 4.53877874702617e-05,
+      "loss": 1.5434,
+      "step": 17458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5247793197631836,
+      "learning_rate": 4.538752312979117e-05,
+      "loss": 1.5403,
+      "step": 17459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9096102714538574,
+      "learning_rate": 4.5387258789320645e-05,
+      "loss": 1.4173,
+      "step": 17460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.386681079864502,
+      "learning_rate": 4.5386994448850126e-05,
+      "loss": 1.2137,
+      "step": 17461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8001372814178467,
+      "learning_rate": 4.538673010837959e-05,
+      "loss": 1.2802,
+      "step": 17462
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8003456592559814,
+      "learning_rate": 4.538646576790907e-05,
+      "loss": 1.3456,
+      "step": 17463
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.191771984100342,
+      "learning_rate": 4.538620142743854e-05,
+      "loss": 1.5096,
+      "step": 17464
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2465922832489014,
+      "learning_rate": 4.538593708696802e-05,
+      "loss": 1.4152,
+      "step": 17465
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7168930768966675,
+      "learning_rate": 4.5385672746497494e-05,
+      "loss": 1.3929,
+      "step": 17466
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.466434359550476,
+      "learning_rate": 4.538540840602696e-05,
+      "loss": 1.3647,
+      "step": 17467
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3627897500991821,
+      "learning_rate": 4.538514406555644e-05,
+      "loss": 1.3011,
+      "step": 17468
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.32684588432312,
+      "learning_rate": 4.538487972508591e-05,
+      "loss": 1.4005,
+      "step": 17469
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7173559665679932,
+      "learning_rate": 4.538461538461539e-05,
+      "loss": 1.418,
+      "step": 17470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.156964063644409,
+      "learning_rate": 4.538435104414486e-05,
+      "loss": 1.5411,
+      "step": 17471
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8696019649505615,
+      "learning_rate": 4.5384086703674335e-05,
+      "loss": 1.3621,
+      "step": 17472
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8760100603103638,
+      "learning_rate": 4.538382236320381e-05,
+      "loss": 1.3128,
+      "step": 17473
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.821537971496582,
+      "learning_rate": 4.538355802273328e-05,
+      "loss": 1.3161,
+      "step": 17474
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9109731912612915,
+      "learning_rate": 4.5383293682262756e-05,
+      "loss": 1.4949,
+      "step": 17475
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0491445064544678,
+      "learning_rate": 4.538302934179223e-05,
+      "loss": 1.3791,
+      "step": 17476
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9898027181625366,
+      "learning_rate": 4.53827650013217e-05,
+      "loss": 1.3721,
+      "step": 17477
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.645795464515686,
+      "learning_rate": 4.5382500660851177e-05,
+      "loss": 1.4293,
+      "step": 17478
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9500731229782104,
+      "learning_rate": 4.538223632038065e-05,
+      "loss": 1.2708,
+      "step": 17479
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.671783447265625,
+      "learning_rate": 4.5381971979910124e-05,
+      "loss": 1.3536,
+      "step": 17480
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6475783586502075,
+      "learning_rate": 4.5381707639439604e-05,
+      "loss": 1.4724,
+      "step": 17481
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.68630850315094,
+      "learning_rate": 4.538144329896907e-05,
+      "loss": 1.4448,
+      "step": 17482
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8327549695968628,
+      "learning_rate": 4.538117895849855e-05,
+      "loss": 1.363,
+      "step": 17483
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.458333969116211,
+      "learning_rate": 4.538091461802802e-05,
+      "loss": 1.4358,
+      "step": 17484
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4466270208358765,
+      "learning_rate": 4.53806502775575e-05,
+      "loss": 1.2207,
+      "step": 17485
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5600645542144775,
+      "learning_rate": 4.538038593708697e-05,
+      "loss": 1.3434,
+      "step": 17486
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.962378740310669,
+      "learning_rate": 4.5380121596616445e-05,
+      "loss": 1.2795,
+      "step": 17487
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9042919874191284,
+      "learning_rate": 4.537985725614592e-05,
+      "loss": 1.4517,
+      "step": 17488
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2525041103363037,
+      "learning_rate": 4.537959291567539e-05,
+      "loss": 1.4144,
+      "step": 17489
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7252248525619507,
+      "learning_rate": 4.5379328575204866e-05,
+      "loss": 1.4853,
+      "step": 17490
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.413888454437256,
+      "learning_rate": 4.537906423473434e-05,
+      "loss": 1.5709,
+      "step": 17491
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6191394329071045,
+      "learning_rate": 4.537879989426381e-05,
+      "loss": 1.6055,
+      "step": 17492
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9709168672561646,
+      "learning_rate": 4.537853555379329e-05,
+      "loss": 1.4644,
+      "step": 17493
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9400017261505127,
+      "learning_rate": 4.537827121332276e-05,
+      "loss": 1.4135,
+      "step": 17494
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.788327693939209,
+      "learning_rate": 4.5378006872852234e-05,
+      "loss": 1.5574,
+      "step": 17495
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.945735216140747,
+      "learning_rate": 4.5377742532381714e-05,
+      "loss": 1.4117,
+      "step": 17496
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3648712635040283,
+      "learning_rate": 4.537747819191118e-05,
+      "loss": 1.5788,
+      "step": 17497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1033082008361816,
+      "learning_rate": 4.537721385144066e-05,
+      "loss": 1.496,
+      "step": 17498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.721322774887085,
+      "learning_rate": 4.537694951097013e-05,
+      "loss": 1.3412,
+      "step": 17499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8425395488739014,
+      "learning_rate": 4.537668517049961e-05,
+      "loss": 1.1596,
+      "step": 17500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9775067567825317,
+      "learning_rate": 4.537642083002908e-05,
+      "loss": 1.4608,
+      "step": 17501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.175001859664917,
+      "learning_rate": 4.537615648955855e-05,
+      "loss": 1.5745,
+      "step": 17502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.601387858390808,
+      "learning_rate": 4.537589214908803e-05,
+      "loss": 1.3013,
+      "step": 17503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6668710708618164,
+      "learning_rate": 4.5375627808617496e-05,
+      "loss": 1.4602,
+      "step": 17504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9177995920181274,
+      "learning_rate": 4.5375363468146976e-05,
+      "loss": 1.2301,
+      "step": 17505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8416507244110107,
+      "learning_rate": 4.537509912767645e-05,
+      "loss": 1.2828,
+      "step": 17506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2672487497329712,
+      "learning_rate": 4.5374834787205924e-05,
+      "loss": 1.1639,
+      "step": 17507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8816543817520142,
+      "learning_rate": 4.53745704467354e-05,
+      "loss": 1.4014,
+      "step": 17508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0774319171905518,
+      "learning_rate": 4.537430610626487e-05,
+      "loss": 1.5076,
+      "step": 17509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.81666100025177,
+      "learning_rate": 4.5374041765794344e-05,
+      "loss": 1.3091,
+      "step": 17510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1746723651885986,
+      "learning_rate": 4.537377742532382e-05,
+      "loss": 1.4867,
+      "step": 17511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9405957460403442,
+      "learning_rate": 4.537351308485329e-05,
+      "loss": 1.4011,
+      "step": 17512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4114623069763184,
+      "learning_rate": 4.5373248744382765e-05,
+      "loss": 1.7385,
+      "step": 17513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7248440980911255,
+      "learning_rate": 4.537298440391224e-05,
+      "loss": 1.4224,
+      "step": 17514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.694357991218567,
+      "learning_rate": 4.537272006344171e-05,
+      "loss": 1.2955,
+      "step": 17515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9975203275680542,
+      "learning_rate": 4.537245572297119e-05,
+      "loss": 1.3927,
+      "step": 17516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7396494150161743,
+      "learning_rate": 4.537219138250066e-05,
+      "loss": 2.2012,
+      "step": 17517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.949212908744812,
+      "learning_rate": 4.537192704203014e-05,
+      "loss": 1.3926,
+      "step": 17518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.352349281311035,
+      "learning_rate": 4.5371662701559606e-05,
+      "loss": 1.4255,
+      "step": 17519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7010440826416016,
+      "learning_rate": 4.537139836108909e-05,
+      "loss": 1.2579,
+      "step": 17520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5682827234268188,
+      "learning_rate": 4.537113402061856e-05,
+      "loss": 1.2843,
+      "step": 17521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7625932693481445,
+      "learning_rate": 4.5370869680148034e-05,
+      "loss": 1.5275,
+      "step": 17522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5977544784545898,
+      "learning_rate": 4.537060533967751e-05,
+      "loss": 1.3278,
+      "step": 17523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.637221336364746,
+      "learning_rate": 4.537034099920698e-05,
+      "loss": 1.3977,
+      "step": 17524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.599855899810791,
+      "learning_rate": 4.5370076658736455e-05,
+      "loss": 1.3498,
+      "step": 17525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5118814706802368,
+      "learning_rate": 4.536981231826593e-05,
+      "loss": 1.34,
+      "step": 17526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5803593397140503,
+      "learning_rate": 4.53695479777954e-05,
+      "loss": 1.3407,
+      "step": 17527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.890571117401123,
+      "learning_rate": 4.5369283637324875e-05,
+      "loss": 1.4979,
+      "step": 17528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.711653709411621,
+      "learning_rate": 4.536901929685435e-05,
+      "loss": 1.4476,
+      "step": 17529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8968255519866943,
+      "learning_rate": 4.536875495638382e-05,
+      "loss": 1.712,
+      "step": 17530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.056654691696167,
+      "learning_rate": 4.53684906159133e-05,
+      "loss": 1.4146,
+      "step": 17531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7801145315170288,
+      "learning_rate": 4.536822627544277e-05,
+      "loss": 1.4145,
+      "step": 17532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8658154010772705,
+      "learning_rate": 4.536796193497225e-05,
+      "loss": 1.4278,
+      "step": 17533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6442837715148926,
+      "learning_rate": 4.536769759450172e-05,
+      "loss": 1.2636,
+      "step": 17534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4187217950820923,
+      "learning_rate": 4.53674332540312e-05,
+      "loss": 1.2616,
+      "step": 17535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6260650157928467,
+      "learning_rate": 4.536716891356067e-05,
+      "loss": 1.3608,
+      "step": 17536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.025315523147583,
+      "learning_rate": 4.5366904573090144e-05,
+      "loss": 1.4778,
+      "step": 17537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.717713475227356,
+      "learning_rate": 4.536664023261962e-05,
+      "loss": 1.4332,
+      "step": 17538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0574069023132324,
+      "learning_rate": 4.5366375892149085e-05,
+      "loss": 1.4044,
+      "step": 17539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9594089984893799,
+      "learning_rate": 4.5366111551678565e-05,
+      "loss": 1.4131,
+      "step": 17540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6724070310592651,
+      "learning_rate": 4.536584721120804e-05,
+      "loss": 1.3288,
+      "step": 17541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8167332410812378,
+      "learning_rate": 4.536558287073751e-05,
+      "loss": 1.4258,
+      "step": 17542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9336743354797363,
+      "learning_rate": 4.5365318530266986e-05,
+      "loss": 1.3873,
+      "step": 17543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.973227858543396,
+      "learning_rate": 4.536505418979646e-05,
+      "loss": 1.5218,
+      "step": 17544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.212390422821045,
+      "learning_rate": 4.536478984932593e-05,
+      "loss": 1.7203,
+      "step": 17545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7108638286590576,
+      "learning_rate": 4.536452550885541e-05,
+      "loss": 1.4121,
+      "step": 17546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.421964406967163,
+      "learning_rate": 4.536426116838488e-05,
+      "loss": 1.5071,
+      "step": 17547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.607282042503357,
+      "learning_rate": 4.5363996827914354e-05,
+      "loss": 1.3764,
+      "step": 17548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6004071235656738,
+      "learning_rate": 4.536373248744383e-05,
+      "loss": 1.3076,
+      "step": 17549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5595585107803345,
+      "learning_rate": 4.53634681469733e-05,
+      "loss": 1.1474,
+      "step": 17550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.809716820716858,
+      "learning_rate": 4.536320380650278e-05,
+      "loss": 1.4074,
+      "step": 17551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.049631118774414,
+      "learning_rate": 4.536293946603225e-05,
+      "loss": 1.4509,
+      "step": 17552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.858546495437622,
+      "learning_rate": 4.536267512556173e-05,
+      "loss": 1.3831,
+      "step": 17553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.986934781074524,
+      "learning_rate": 4.5362410785091195e-05,
+      "loss": 1.5368,
+      "step": 17554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6507905721664429,
+      "learning_rate": 4.5362146444620675e-05,
+      "loss": 1.3284,
+      "step": 17555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4803040027618408,
+      "learning_rate": 4.536188210415015e-05,
+      "loss": 1.2409,
+      "step": 17556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.326481580734253,
+      "learning_rate": 4.536161776367962e-05,
+      "loss": 1.4763,
+      "step": 17557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9819523096084595,
+      "learning_rate": 4.5361353423209096e-05,
+      "loss": 1.4924,
+      "step": 17558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7983555793762207,
+      "learning_rate": 4.536108908273857e-05,
+      "loss": 1.3789,
+      "step": 17559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1035492420196533,
+      "learning_rate": 4.536082474226804e-05,
+      "loss": 1.3705,
+      "step": 17560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8553069829940796,
+      "learning_rate": 4.536056040179752e-05,
+      "loss": 2.3047,
+      "step": 17561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2807940244674683,
+      "learning_rate": 4.536029606132699e-05,
+      "loss": 1.3561,
+      "step": 17562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1950786113739014,
+      "learning_rate": 4.5360031720856464e-05,
+      "loss": 1.4939,
+      "step": 17563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9409102201461792,
+      "learning_rate": 4.535976738038594e-05,
+      "loss": 1.3461,
+      "step": 17564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6370652914047241,
+      "learning_rate": 4.535950303991541e-05,
+      "loss": 1.3794,
+      "step": 17565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.051654100418091,
+      "learning_rate": 4.535923869944489e-05,
+      "loss": 1.3979,
+      "step": 17566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8698750734329224,
+      "learning_rate": 4.535897435897436e-05,
+      "loss": 1.6027,
+      "step": 17567
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3887290954589844,
+      "learning_rate": 4.535871001850384e-05,
+      "loss": 1.4071,
+      "step": 17568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5108050107955933,
+      "learning_rate": 4.5358445678033305e-05,
+      "loss": 1.2602,
+      "step": 17569
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6783862113952637,
+      "learning_rate": 4.5358181337562786e-05,
+      "loss": 1.3801,
+      "step": 17570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.576885223388672,
+      "learning_rate": 4.535791699709226e-05,
+      "loss": 1.4404,
+      "step": 17571
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5159765481948853,
+      "learning_rate": 4.535765265662173e-05,
+      "loss": 1.2246,
+      "step": 17572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4777610301971436,
+      "learning_rate": 4.5357388316151206e-05,
+      "loss": 1.3951,
+      "step": 17573
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6784077882766724,
+      "learning_rate": 4.535712397568067e-05,
+      "loss": 1.394,
+      "step": 17574
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4171266555786133,
+      "learning_rate": 4.5356859635210154e-05,
+      "loss": 1.6354,
+      "step": 17575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2750139236450195,
+      "learning_rate": 4.535659529473963e-05,
+      "loss": 1.0692,
+      "step": 17576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0968456268310547,
+      "learning_rate": 4.53563309542691e-05,
+      "loss": 1.4988,
+      "step": 17577
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7462040185928345,
+      "learning_rate": 4.5356066613798574e-05,
+      "loss": 1.2797,
+      "step": 17578
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1941306591033936,
+      "learning_rate": 4.535580227332805e-05,
+      "loss": 1.5699,
+      "step": 17579
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.610113620758057,
+      "learning_rate": 4.535553793285752e-05,
+      "loss": 1.4083,
+      "step": 17580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5793241262435913,
+      "learning_rate": 4.5355273592387e-05,
+      "loss": 1.2627,
+      "step": 17581
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1194562911987305,
+      "learning_rate": 4.535500925191647e-05,
+      "loss": 1.5234,
+      "step": 17582
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.562255859375,
+      "learning_rate": 4.535474491144595e-05,
+      "loss": 1.4073,
+      "step": 17583
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3407659530639648,
+      "learning_rate": 4.5354480570975416e-05,
+      "loss": 1.8514,
+      "step": 17584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9921233654022217,
+      "learning_rate": 4.535421623050489e-05,
+      "loss": 1.2806,
+      "step": 17585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3608077764511108,
+      "learning_rate": 4.535395189003437e-05,
+      "loss": 1.2713,
+      "step": 17586
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6842963695526123,
+      "learning_rate": 4.5353687549563836e-05,
+      "loss": 1.5952,
+      "step": 17587
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8893866539001465,
+      "learning_rate": 4.535342320909332e-05,
+      "loss": 1.434,
+      "step": 17588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8375954627990723,
+      "learning_rate": 4.5353158868622784e-05,
+      "loss": 1.3397,
+      "step": 17589
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.715690016746521,
+      "learning_rate": 4.5352894528152264e-05,
+      "loss": 1.3348,
+      "step": 17590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7611085176467896,
+      "learning_rate": 4.535263018768174e-05,
+      "loss": 1.4382,
+      "step": 17591
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7648038864135742,
+      "learning_rate": 4.535236584721121e-05,
+      "loss": 1.4215,
+      "step": 17592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9047482013702393,
+      "learning_rate": 4.5352101506740685e-05,
+      "loss": 1.4296,
+      "step": 17593
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.658172845840454,
+      "learning_rate": 4.535183716627016e-05,
+      "loss": 1.5777,
+      "step": 17594
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8741570711135864,
+      "learning_rate": 4.535157282579963e-05,
+      "loss": 1.5644,
+      "step": 17595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.357423782348633,
+      "learning_rate": 4.5351308485329105e-05,
+      "loss": 1.5496,
+      "step": 17596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5231738090515137,
+      "learning_rate": 4.535104414485858e-05,
+      "loss": 1.3107,
+      "step": 17597
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4560952186584473,
+      "learning_rate": 4.535077980438805e-05,
+      "loss": 1.3352,
+      "step": 17598
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7034807205200195,
+      "learning_rate": 4.5350515463917526e-05,
+      "loss": 1.3003,
+      "step": 17599
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0479824542999268,
+      "learning_rate": 4.5350251123447e-05,
+      "loss": 1.4877,
+      "step": 17600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5134280920028687,
+      "learning_rate": 4.534998678297648e-05,
+      "loss": 1.3429,
+      "step": 17601
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.597355842590332,
+      "learning_rate": 4.534972244250595e-05,
+      "loss": 1.2637,
+      "step": 17602
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.665988564491272,
+      "learning_rate": 4.534945810203543e-05,
+      "loss": 1.4008,
+      "step": 17603
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.632755994796753,
+      "learning_rate": 4.5349193761564894e-05,
+      "loss": 1.5957,
+      "step": 17604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.101200580596924,
+      "learning_rate": 4.5348929421094374e-05,
+      "loss": 1.3006,
+      "step": 17605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6785489320755005,
+      "learning_rate": 4.534866508062385e-05,
+      "loss": 1.2191,
+      "step": 17606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6219733953475952,
+      "learning_rate": 4.534840074015332e-05,
+      "loss": 1.4263,
+      "step": 17607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4257620573043823,
+      "learning_rate": 4.5348136399682795e-05,
+      "loss": 1.7808,
+      "step": 17608
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.87204909324646,
+      "learning_rate": 4.534787205921227e-05,
+      "loss": 1.3659,
+      "step": 17609
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8603264093399048,
+      "learning_rate": 4.534760771874174e-05,
+      "loss": 1.2627,
+      "step": 17610
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9249823093414307,
+      "learning_rate": 4.5347343378271216e-05,
+      "loss": 1.5724,
+      "step": 17611
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8524645566940308,
+      "learning_rate": 4.534707903780069e-05,
+      "loss": 1.3244,
+      "step": 17612
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9497523307800293,
+      "learning_rate": 4.534681469733016e-05,
+      "loss": 1.326,
+      "step": 17613
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2612926959991455,
+      "learning_rate": 4.5346550356859636e-05,
+      "loss": 1.6117,
+      "step": 17614
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.455092430114746,
+      "learning_rate": 4.534628601638911e-05,
+      "loss": 1.6763,
+      "step": 17615
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0597758293151855,
+      "learning_rate": 4.534602167591859e-05,
+      "loss": 1.3615,
+      "step": 17616
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9647506475448608,
+      "learning_rate": 4.534575733544806e-05,
+      "loss": 1.3799,
+      "step": 17617
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6299771070480347,
+      "learning_rate": 4.534549299497754e-05,
+      "loss": 1.2473,
+      "step": 17618
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.531542181968689,
+      "learning_rate": 4.5345228654507004e-05,
+      "loss": 1.4519,
+      "step": 17619
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8685886859893799,
+      "learning_rate": 4.534496431403648e-05,
+      "loss": 1.2267,
+      "step": 17620
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0446486473083496,
+      "learning_rate": 4.534469997356596e-05,
+      "loss": 1.9027,
+      "step": 17621
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0991482734680176,
+      "learning_rate": 4.5344435633095425e-05,
+      "loss": 1.2042,
+      "step": 17622
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.914549708366394,
+      "learning_rate": 4.5344171292624905e-05,
+      "loss": 1.5004,
+      "step": 17623
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.707027792930603,
+      "learning_rate": 4.534390695215437e-05,
+      "loss": 1.4658,
+      "step": 17624
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5700792074203491,
+      "learning_rate": 4.534364261168385e-05,
+      "loss": 1.4357,
+      "step": 17625
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.843729853630066,
+      "learning_rate": 4.5343378271213326e-05,
+      "loss": 1.3117,
+      "step": 17626
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.002254009246826,
+      "learning_rate": 4.53431139307428e-05,
+      "loss": 1.5155,
+      "step": 17627
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6288691759109497,
+      "learning_rate": 4.534284959027227e-05,
+      "loss": 1.4294,
+      "step": 17628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5655274391174316,
+      "learning_rate": 4.534258524980175e-05,
+      "loss": 1.3251,
+      "step": 17629
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7560862302780151,
+      "learning_rate": 4.534232090933122e-05,
+      "loss": 1.4084,
+      "step": 17630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6200166940689087,
+      "learning_rate": 4.5342056568860694e-05,
+      "loss": 1.3916,
+      "step": 17631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.712883710861206,
+      "learning_rate": 4.534179222839017e-05,
+      "loss": 1.502,
+      "step": 17632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3963596820831299,
+      "learning_rate": 4.534152788791964e-05,
+      "loss": 1.3564,
+      "step": 17633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0483803749084473,
+      "learning_rate": 4.5341263547449115e-05,
+      "loss": 1.4855,
+      "step": 17634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3908064365386963,
+      "learning_rate": 4.534099920697859e-05,
+      "loss": 1.4052,
+      "step": 17635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7454509735107422,
+      "learning_rate": 4.534073486650807e-05,
+      "loss": 1.4065,
+      "step": 17636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.867102026939392,
+      "learning_rate": 4.5340470526037535e-05,
+      "loss": 1.4801,
+      "step": 17637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6402788162231445,
+      "learning_rate": 4.5340206185567016e-05,
+      "loss": 1.5816,
+      "step": 17638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.821161270141602,
+      "learning_rate": 4.533994184509648e-05,
+      "loss": 1.5924,
+      "step": 17639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5169116258621216,
+      "learning_rate": 4.533967750462596e-05,
+      "loss": 1.3039,
+      "step": 17640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7477515935897827,
+      "learning_rate": 4.5339413164155436e-05,
+      "loss": 1.4093,
+      "step": 17641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6299989223480225,
+      "learning_rate": 4.533914882368491e-05,
+      "loss": 1.4224,
+      "step": 17642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1089510917663574,
+      "learning_rate": 4.5338884483214383e-05,
+      "loss": 1.3358,
+      "step": 17643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.625569462776184,
+      "learning_rate": 4.533862014274386e-05,
+      "loss": 1.4861,
+      "step": 17644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7284401655197144,
+      "learning_rate": 4.533835580227333e-05,
+      "loss": 2.4107,
+      "step": 17645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4493989944458008,
+      "learning_rate": 4.5338091461802804e-05,
+      "loss": 1.191,
+      "step": 17646
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3437910079956055,
+      "learning_rate": 4.533782712133228e-05,
+      "loss": 1.3049,
+      "step": 17647
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6976124048233032,
+      "learning_rate": 4.533756278086175e-05,
+      "loss": 1.2012,
+      "step": 17648
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6651318073272705,
+      "learning_rate": 4.5337298440391225e-05,
+      "loss": 1.445,
+      "step": 17649
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3010125160217285,
+      "learning_rate": 4.53370340999207e-05,
+      "loss": 1.5966,
+      "step": 17650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.995129942893982,
+      "learning_rate": 4.533676975945018e-05,
+      "loss": 1.3816,
+      "step": 17651
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9282522201538086,
+      "learning_rate": 4.5336505418979646e-05,
+      "loss": 1.35,
+      "step": 17652
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.561785936355591,
+      "learning_rate": 4.5336241078509126e-05,
+      "loss": 2.0943,
+      "step": 17653
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1638123989105225,
+      "learning_rate": 4.533597673803859e-05,
+      "loss": 1.3615,
+      "step": 17654
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3430540561676025,
+      "learning_rate": 4.533571239756807e-05,
+      "loss": 2.1206,
+      "step": 17655
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5952513217926025,
+      "learning_rate": 4.533544805709755e-05,
+      "loss": 1.2851,
+      "step": 17656
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.837026119232178,
+      "learning_rate": 4.5335183716627013e-05,
+      "loss": 1.6864,
+      "step": 17657
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9426345825195312,
+      "learning_rate": 4.5334919376156494e-05,
+      "loss": 1.3507,
+      "step": 17658
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9145216941833496,
+      "learning_rate": 4.533465503568596e-05,
+      "loss": 1.3631,
+      "step": 17659
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.031140089035034,
+      "learning_rate": 4.533439069521544e-05,
+      "loss": 2.0152,
+      "step": 17660
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5592604875564575,
+      "learning_rate": 4.5334126354744915e-05,
+      "loss": 1.358,
+      "step": 17661
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4544116258621216,
+      "learning_rate": 4.533386201427439e-05,
+      "loss": 1.3359,
+      "step": 17662
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6207897663116455,
+      "learning_rate": 4.533359767380386e-05,
+      "loss": 1.2981,
+      "step": 17663
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.504875659942627,
+      "learning_rate": 4.5333333333333335e-05,
+      "loss": 1.4665,
+      "step": 17664
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9316869974136353,
+      "learning_rate": 4.533306899286281e-05,
+      "loss": 1.3701,
+      "step": 17665
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1709632873535156,
+      "learning_rate": 4.533280465239228e-05,
+      "loss": 1.4074,
+      "step": 17666
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5705479383468628,
+      "learning_rate": 4.5332540311921756e-05,
+      "loss": 1.3278,
+      "step": 17667
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.003305673599243,
+      "learning_rate": 4.533227597145123e-05,
+      "loss": 1.3791,
+      "step": 17668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7388556003570557,
+      "learning_rate": 4.53320116309807e-05,
+      "loss": 1.771,
+      "step": 17669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4091076850891113,
+      "learning_rate": 4.533174729051018e-05,
+      "loss": 1.4247,
+      "step": 17670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.055133819580078,
+      "learning_rate": 4.533148295003966e-05,
+      "loss": 1.3136,
+      "step": 17671
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4923174381256104,
+      "learning_rate": 4.5331218609569124e-05,
+      "loss": 1.3289,
+      "step": 17672
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6851072311401367,
+      "learning_rate": 4.5330954269098604e-05,
+      "loss": 1.4405,
+      "step": 17673
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7361177206039429,
+      "learning_rate": 4.533068992862807e-05,
+      "loss": 1.1992,
+      "step": 17674
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.833153247833252,
+      "learning_rate": 4.533042558815755e-05,
+      "loss": 1.3816,
+      "step": 17675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8543033599853516,
+      "learning_rate": 4.5330161247687025e-05,
+      "loss": 1.5435,
+      "step": 17676
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4812798500061035,
+      "learning_rate": 4.53298969072165e-05,
+      "loss": 1.4842,
+      "step": 17677
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6211059093475342,
+      "learning_rate": 4.532963256674597e-05,
+      "loss": 1.1837,
+      "step": 17678
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7874587774276733,
+      "learning_rate": 4.5329368226275446e-05,
+      "loss": 1.4812,
+      "step": 17679
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.449650764465332,
+      "learning_rate": 4.532910388580492e-05,
+      "loss": 1.4082,
+      "step": 17680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5665254592895508,
+      "learning_rate": 4.532883954533439e-05,
+      "loss": 1.3812,
+      "step": 17681
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4049248695373535,
+      "learning_rate": 4.5328575204863866e-05,
+      "loss": 1.5211,
+      "step": 17682
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2138993740081787,
+      "learning_rate": 4.532831086439334e-05,
+      "loss": 1.3174,
+      "step": 17683
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.952813744544983,
+      "learning_rate": 4.5328046523922813e-05,
+      "loss": 1.362,
+      "step": 17684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5719127655029297,
+      "learning_rate": 4.532778218345229e-05,
+      "loss": 1.4491,
+      "step": 17685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8656131029129028,
+      "learning_rate": 4.532751784298177e-05,
+      "loss": 1.5731,
+      "step": 17686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7817734479904175,
+      "learning_rate": 4.5327253502511234e-05,
+      "loss": 1.5536,
+      "step": 17687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6260524988174438,
+      "learning_rate": 4.5326989162040714e-05,
+      "loss": 1.5064,
+      "step": 17688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2987213134765625,
+      "learning_rate": 4.532672482157018e-05,
+      "loss": 1.5816,
+      "step": 17689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0625972747802734,
+      "learning_rate": 4.532646048109966e-05,
+      "loss": 1.4177,
+      "step": 17690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8368161916732788,
+      "learning_rate": 4.5326196140629135e-05,
+      "loss": 1.3536,
+      "step": 17691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5417248010635376,
+      "learning_rate": 4.53259318001586e-05,
+      "loss": 1.4582,
+      "step": 17692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.622833013534546,
+      "learning_rate": 4.532566745968808e-05,
+      "loss": 1.4987,
+      "step": 17693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9372923374176025,
+      "learning_rate": 4.532540311921755e-05,
+      "loss": 1.3709,
+      "step": 17694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9712944030761719,
+      "learning_rate": 4.532513877874703e-05,
+      "loss": 1.5201,
+      "step": 17695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9820588827133179,
+      "learning_rate": 4.53248744382765e-05,
+      "loss": 1.4237,
+      "step": 17696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.556483507156372,
+      "learning_rate": 4.532461009780598e-05,
+      "loss": 1.3686,
+      "step": 17697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5779796838760376,
+      "learning_rate": 4.532434575733545e-05,
+      "loss": 1.3894,
+      "step": 17698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5693167448043823,
+      "learning_rate": 4.5324081416864924e-05,
+      "loss": 1.2193,
+      "step": 17699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4285756349563599,
+      "learning_rate": 4.53238170763944e-05,
+      "loss": 1.2523,
+      "step": 17700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9322178363800049,
+      "learning_rate": 4.532355273592387e-05,
+      "loss": 1.4252,
+      "step": 17701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.42444908618927,
+      "learning_rate": 4.5323288395453344e-05,
+      "loss": 1.3494,
+      "step": 17702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.033130407333374,
+      "learning_rate": 4.532302405498282e-05,
+      "loss": 1.4952,
+      "step": 17703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8025693893432617,
+      "learning_rate": 4.532275971451229e-05,
+      "loss": 1.3556,
+      "step": 17704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2894853353500366,
+      "learning_rate": 4.5322495374041765e-05,
+      "loss": 1.3409,
+      "step": 17705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9280809164047241,
+      "learning_rate": 4.5322231033571246e-05,
+      "loss": 1.4286,
+      "step": 17706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9258732795715332,
+      "learning_rate": 4.532196669310071e-05,
+      "loss": 1.3976,
+      "step": 17707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4327025413513184,
+      "learning_rate": 4.532170235263019e-05,
+      "loss": 1.4497,
+      "step": 17708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9996522665023804,
+      "learning_rate": 4.532143801215966e-05,
+      "loss": 1.5283,
+      "step": 17709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6822149753570557,
+      "learning_rate": 4.532117367168914e-05,
+      "loss": 1.3617,
+      "step": 17710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2593441009521484,
+      "learning_rate": 4.5320909331218613e-05,
+      "loss": 1.5041,
+      "step": 17711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.910634994506836,
+      "learning_rate": 4.532064499074809e-05,
+      "loss": 1.354,
+      "step": 17712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8411364555358887,
+      "learning_rate": 4.532038065027756e-05,
+      "loss": 1.4362,
+      "step": 17713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1493916511535645,
+      "learning_rate": 4.5320116309807034e-05,
+      "loss": 1.3789,
+      "step": 17714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5401980876922607,
+      "learning_rate": 4.531985196933651e-05,
+      "loss": 1.1809,
+      "step": 17715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3778836727142334,
+      "learning_rate": 4.531958762886598e-05,
+      "loss": 1.4742,
+      "step": 17716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6897799968719482,
+      "learning_rate": 4.5319323288395455e-05,
+      "loss": 1.3474,
+      "step": 17717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6295650005340576,
+      "learning_rate": 4.531905894792493e-05,
+      "loss": 1.3338,
+      "step": 17718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.623224973678589,
+      "learning_rate": 4.53187946074544e-05,
+      "loss": 1.3655,
+      "step": 17719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.924332618713379,
+      "learning_rate": 4.5318530266983876e-05,
+      "loss": 1.3358,
+      "step": 17720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5546603202819824,
+      "learning_rate": 4.5318265926513356e-05,
+      "loss": 1.3397,
+      "step": 17721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.518918514251709,
+      "learning_rate": 4.531800158604282e-05,
+      "loss": 1.3837,
+      "step": 17722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.418814182281494,
+      "learning_rate": 4.53177372455723e-05,
+      "loss": 1.3949,
+      "step": 17723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6310077905654907,
+      "learning_rate": 4.531747290510177e-05,
+      "loss": 1.3811,
+      "step": 17724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9183909893035889,
+      "learning_rate": 4.531720856463125e-05,
+      "loss": 1.4513,
+      "step": 17725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3657257556915283,
+      "learning_rate": 4.5316944224160724e-05,
+      "loss": 1.687,
+      "step": 17726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4474289417266846,
+      "learning_rate": 4.53166798836902e-05,
+      "loss": 1.5325,
+      "step": 17727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9023419618606567,
+      "learning_rate": 4.531641554321967e-05,
+      "loss": 1.3163,
+      "step": 17728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.544097661972046,
+      "learning_rate": 4.531615120274914e-05,
+      "loss": 1.4655,
+      "step": 17729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.584178924560547,
+      "learning_rate": 4.531588686227862e-05,
+      "loss": 1.4534,
+      "step": 17730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5500731468200684,
+      "learning_rate": 4.531562252180809e-05,
+      "loss": 1.4978,
+      "step": 17731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.587952733039856,
+      "learning_rate": 4.5315358181337565e-05,
+      "loss": 1.4026,
+      "step": 17732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6184614896774292,
+      "learning_rate": 4.531509384086704e-05,
+      "loss": 1.3327,
+      "step": 17733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8561784029006958,
+      "learning_rate": 4.531482950039651e-05,
+      "loss": 1.4,
+      "step": 17734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3184709548950195,
+      "learning_rate": 4.5314565159925986e-05,
+      "loss": 1.159,
+      "step": 17735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9659688472747803,
+      "learning_rate": 4.5314300819455466e-05,
+      "loss": 1.5042,
+      "step": 17736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6546945571899414,
+      "learning_rate": 4.531403647898493e-05,
+      "loss": 1.3326,
+      "step": 17737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6040146350860596,
+      "learning_rate": 4.5313772138514407e-05,
+      "loss": 1.3215,
+      "step": 17738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.819894552230835,
+      "learning_rate": 4.531350779804388e-05,
+      "loss": 1.5395,
+      "step": 17739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.891502857208252,
+      "learning_rate": 4.5313243457573354e-05,
+      "loss": 1.2871,
+      "step": 17740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7433905601501465,
+      "learning_rate": 4.5312979117102834e-05,
+      "loss": 1.3503,
+      "step": 17741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6970679759979248,
+      "learning_rate": 4.53127147766323e-05,
+      "loss": 1.3547,
+      "step": 17742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6501531600952148,
+      "learning_rate": 4.531245043616178e-05,
+      "loss": 1.3642,
+      "step": 17743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7120907306671143,
+      "learning_rate": 4.531218609569125e-05,
+      "loss": 1.3956,
+      "step": 17744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9153589010238647,
+      "learning_rate": 4.531192175522073e-05,
+      "loss": 2.2104,
+      "step": 17745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0206832885742188,
+      "learning_rate": 4.53116574147502e-05,
+      "loss": 1.3188,
+      "step": 17746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.095996379852295,
+      "learning_rate": 4.5311393074279676e-05,
+      "loss": 1.4288,
+      "step": 17747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4972633123397827,
+      "learning_rate": 4.531112873380915e-05,
+      "loss": 1.3822,
+      "step": 17748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0969326496124268,
+      "learning_rate": 4.531086439333862e-05,
+      "loss": 1.395,
+      "step": 17749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6617708206176758,
+      "learning_rate": 4.5310600052868096e-05,
+      "loss": 1.4039,
+      "step": 17750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0509798526763916,
+      "learning_rate": 4.531033571239757e-05,
+      "loss": 1.6319,
+      "step": 17751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6194757223129272,
+      "learning_rate": 4.531007137192704e-05,
+      "loss": 1.2174,
+      "step": 17752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.052278995513916,
+      "learning_rate": 4.530980703145652e-05,
+      "loss": 1.4952,
+      "step": 17753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6141316890716553,
+      "learning_rate": 4.530954269098599e-05,
+      "loss": 2.2094,
+      "step": 17754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6630574464797974,
+      "learning_rate": 4.5309278350515464e-05,
+      "loss": 1.2582,
+      "step": 17755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5225507020950317,
+      "learning_rate": 4.5309014010044944e-05,
+      "loss": 1.2795,
+      "step": 17756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9484508037567139,
+      "learning_rate": 4.530874966957441e-05,
+      "loss": 1.3643,
+      "step": 17757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6512477397918701,
+      "learning_rate": 4.530848532910389e-05,
+      "loss": 1.4374,
+      "step": 17758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6065075397491455,
+      "learning_rate": 4.530822098863336e-05,
+      "loss": 1.3953,
+      "step": 17759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.060447931289673,
+      "learning_rate": 4.530795664816284e-05,
+      "loss": 1.4842,
+      "step": 17760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5164517164230347,
+      "learning_rate": 4.530769230769231e-05,
+      "loss": 1.4477,
+      "step": 17761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4702134132385254,
+      "learning_rate": 4.5307427967221786e-05,
+      "loss": 1.4588,
+      "step": 17762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3449862003326416,
+      "learning_rate": 4.530716362675126e-05,
+      "loss": 1.1676,
+      "step": 17763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6291741132736206,
+      "learning_rate": 4.5306899286280726e-05,
+      "loss": 1.4986,
+      "step": 17764
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.173649787902832,
+      "learning_rate": 4.5306634945810207e-05,
+      "loss": 1.3877,
+      "step": 17765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7874724864959717,
+      "learning_rate": 4.530637060533968e-05,
+      "loss": 1.3925,
+      "step": 17766
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9587939977645874,
+      "learning_rate": 4.5306106264869154e-05,
+      "loss": 1.4914,
+      "step": 17767
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8688790798187256,
+      "learning_rate": 4.530584192439863e-05,
+      "loss": 1.3917,
+      "step": 17768
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7224162817001343,
+      "learning_rate": 4.53055775839281e-05,
+      "loss": 1.2652,
+      "step": 17769
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9454283714294434,
+      "learning_rate": 4.5305313243457574e-05,
+      "loss": 1.4068,
+      "step": 17770
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.049480438232422,
+      "learning_rate": 4.5305048902987055e-05,
+      "loss": 1.6063,
+      "step": 17771
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.457226276397705,
+      "learning_rate": 4.530478456251652e-05,
+      "loss": 1.4724,
+      "step": 17772
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6467629671096802,
+      "learning_rate": 4.5304520222046e-05,
+      "loss": 1.4056,
+      "step": 17773
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1387250423431396,
+      "learning_rate": 4.530425588157547e-05,
+      "loss": 1.387,
+      "step": 17774
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9217005968093872,
+      "learning_rate": 4.530399154110494e-05,
+      "loss": 1.231,
+      "step": 17775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9632766246795654,
+      "learning_rate": 4.530372720063442e-05,
+      "loss": 1.4776,
+      "step": 17776
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9419009685516357,
+      "learning_rate": 4.530346286016389e-05,
+      "loss": 1.5492,
+      "step": 17777
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8242988586425781,
+      "learning_rate": 4.530319851969337e-05,
+      "loss": 1.4954,
+      "step": 17778
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.533254623413086,
+      "learning_rate": 4.5302934179222837e-05,
+      "loss": 1.2772,
+      "step": 17779
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7516146898269653,
+      "learning_rate": 4.530266983875232e-05,
+      "loss": 1.4472,
+      "step": 17780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7646384239196777,
+      "learning_rate": 4.530240549828179e-05,
+      "loss": 1.279,
+      "step": 17781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.557986855506897,
+      "learning_rate": 4.5302141157811264e-05,
+      "loss": 1.3396,
+      "step": 17782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.211712598800659,
+      "learning_rate": 4.530187681734074e-05,
+      "loss": 1.5513,
+      "step": 17783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.853369951248169,
+      "learning_rate": 4.530161247687021e-05,
+      "loss": 1.5169,
+      "step": 17784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8775817155838013,
+      "learning_rate": 4.5301348136399685e-05,
+      "loss": 1.4267,
+      "step": 17785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2514376640319824,
+      "learning_rate": 4.530108379592916e-05,
+      "loss": 1.4128,
+      "step": 17786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.527262568473816,
+      "learning_rate": 4.530081945545863e-05,
+      "loss": 1.3253,
+      "step": 17787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.51814866065979,
+      "learning_rate": 4.5300555114988105e-05,
+      "loss": 1.3487,
+      "step": 17788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5553128719329834,
+      "learning_rate": 4.530029077451758e-05,
+      "loss": 1.3289,
+      "step": 17789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0280778408050537,
+      "learning_rate": 4.530002643404705e-05,
+      "loss": 1.417,
+      "step": 17790
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2303872108459473,
+      "learning_rate": 4.529976209357653e-05,
+      "loss": 1.4501,
+      "step": 17791
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2997090816497803,
+      "learning_rate": 4.5299497753106e-05,
+      "loss": 1.6372,
+      "step": 17792
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1840128898620605,
+      "learning_rate": 4.529923341263548e-05,
+      "loss": 1.4076,
+      "step": 17793
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9449145793914795,
+      "learning_rate": 4.529896907216495e-05,
+      "loss": 1.5122,
+      "step": 17794
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8156423568725586,
+      "learning_rate": 4.529870473169443e-05,
+      "loss": 1.3629,
+      "step": 17795
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0954456329345703,
+      "learning_rate": 4.52984403912239e-05,
+      "loss": 1.6701,
+      "step": 17796
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3599283695220947,
+      "learning_rate": 4.5298176050753374e-05,
+      "loss": 1.3305,
+      "step": 17797
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6643935441970825,
+      "learning_rate": 4.529791171028285e-05,
+      "loss": 1.2253,
+      "step": 17798
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9412797689437866,
+      "learning_rate": 4.529764736981232e-05,
+      "loss": 1.3412,
+      "step": 17799
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5781519412994385,
+      "learning_rate": 4.5297383029341795e-05,
+      "loss": 1.3449,
+      "step": 17800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.723083734512329,
+      "learning_rate": 4.529711868887127e-05,
+      "loss": 1.3208,
+      "step": 17801
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9399280548095703,
+      "learning_rate": 4.529685434840074e-05,
+      "loss": 1.6797,
+      "step": 17802
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9419262409210205,
+      "learning_rate": 4.5296590007930216e-05,
+      "loss": 1.1747,
+      "step": 17803
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.747104525566101,
+      "learning_rate": 4.529632566745969e-05,
+      "loss": 1.5015,
+      "step": 17804
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0735790729522705,
+      "learning_rate": 4.529606132698916e-05,
+      "loss": 2.1602,
+      "step": 17805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6756818294525146,
+      "learning_rate": 4.529579698651864e-05,
+      "loss": 1.228,
+      "step": 17806
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6628923416137695,
+      "learning_rate": 4.529553264604811e-05,
+      "loss": 1.3379,
+      "step": 17807
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4110859632492065,
+      "learning_rate": 4.529526830557759e-05,
+      "loss": 1.6896,
+      "step": 17808
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2280144691467285,
+      "learning_rate": 4.529500396510706e-05,
+      "loss": 1.2831,
+      "step": 17809
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4168888330459595,
+      "learning_rate": 4.529473962463653e-05,
+      "loss": 1.3804,
+      "step": 17810
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8922009468078613,
+      "learning_rate": 4.529447528416601e-05,
+      "loss": 1.6242,
+      "step": 17811
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5401607751846313,
+      "learning_rate": 4.529421094369548e-05,
+      "loss": 1.2871,
+      "step": 17812
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6697053909301758,
+      "learning_rate": 4.529394660322496e-05,
+      "loss": 1.8128,
+      "step": 17813
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4686710834503174,
+      "learning_rate": 4.5293682262754425e-05,
+      "loss": 1.32,
+      "step": 17814
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8074491024017334,
+      "learning_rate": 4.5293417922283905e-05,
+      "loss": 1.5367,
+      "step": 17815
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.716341495513916,
+      "learning_rate": 4.529315358181338e-05,
+      "loss": 1.4679,
+      "step": 17816
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.606270432472229,
+      "learning_rate": 4.529288924134285e-05,
+      "loss": 1.2972,
+      "step": 17817
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.697368621826172,
+      "learning_rate": 4.5292624900872326e-05,
+      "loss": 1.5311,
+      "step": 17818
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5176265239715576,
+      "learning_rate": 4.52923605604018e-05,
+      "loss": 1.4415,
+      "step": 17819
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7218525409698486,
+      "learning_rate": 4.529209621993127e-05,
+      "loss": 1.3804,
+      "step": 17820
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7554326057434082,
+      "learning_rate": 4.529183187946075e-05,
+      "loss": 1.5096,
+      "step": 17821
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2269070148468018,
+      "learning_rate": 4.529156753899022e-05,
+      "loss": 1.3279,
+      "step": 17822
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6378295421600342,
+      "learning_rate": 4.5291303198519694e-05,
+      "loss": 1.4456,
+      "step": 17823
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7688632011413574,
+      "learning_rate": 4.529103885804917e-05,
+      "loss": 1.4785,
+      "step": 17824
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1172285079956055,
+      "learning_rate": 4.529077451757864e-05,
+      "loss": 1.5618,
+      "step": 17825
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9793368577957153,
+      "learning_rate": 4.529051017710812e-05,
+      "loss": 1.4076,
+      "step": 17826
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5318599939346313,
+      "learning_rate": 4.529024583663759e-05,
+      "loss": 1.4092,
+      "step": 17827
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4364967346191406,
+      "learning_rate": 4.528998149616707e-05,
+      "loss": 1.3434,
+      "step": 17828
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0166616439819336,
+      "learning_rate": 4.5289717155696535e-05,
+      "loss": 1.3927,
+      "step": 17829
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.262070655822754,
+      "learning_rate": 4.5289452815226016e-05,
+      "loss": 1.4609,
+      "step": 17830
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6395124197006226,
+      "learning_rate": 4.528918847475549e-05,
+      "loss": 1.4667,
+      "step": 17831
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6198740005493164,
+      "learning_rate": 4.528892413428496e-05,
+      "loss": 1.4048,
+      "step": 17832
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5140478610992432,
+      "learning_rate": 4.5288659793814436e-05,
+      "loss": 1.3001,
+      "step": 17833
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.744055986404419,
+      "learning_rate": 4.528839545334391e-05,
+      "loss": 1.4553,
+      "step": 17834
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9501752853393555,
+      "learning_rate": 4.5288131112873384e-05,
+      "loss": 1.4498,
+      "step": 17835
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.070188283920288,
+      "learning_rate": 4.528786677240286e-05,
+      "loss": 1.3901,
+      "step": 17836
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7237995862960815,
+      "learning_rate": 4.528760243193233e-05,
+      "loss": 1.5284,
+      "step": 17837
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8454148769378662,
+      "learning_rate": 4.5287338091461804e-05,
+      "loss": 1.7651,
+      "step": 17838
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.710524082183838,
+      "learning_rate": 4.528707375099128e-05,
+      "loss": 1.327,
+      "step": 17839
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.030729055404663,
+      "learning_rate": 4.528680941052075e-05,
+      "loss": 1.66,
+      "step": 17840
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6308156251907349,
+      "learning_rate": 4.528654507005023e-05,
+      "loss": 1.198,
+      "step": 17841
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.141331672668457,
+      "learning_rate": 4.52862807295797e-05,
+      "loss": 1.3572,
+      "step": 17842
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.368173599243164,
+      "learning_rate": 4.528601638910918e-05,
+      "loss": 1.2329,
+      "step": 17843
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6030278205871582,
+      "learning_rate": 4.5285752048638646e-05,
+      "loss": 1.2124,
+      "step": 17844
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6906700134277344,
+      "learning_rate": 4.5285487708168126e-05,
+      "loss": 1.2395,
+      "step": 17845
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.584509015083313,
+      "learning_rate": 4.52852233676976e-05,
+      "loss": 1.354,
+      "step": 17846
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.620863437652588,
+      "learning_rate": 4.5284959027227066e-05,
+      "loss": 1.2961,
+      "step": 17847
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6886979341506958,
+      "learning_rate": 4.528469468675655e-05,
+      "loss": 1.383,
+      "step": 17848
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0780446529388428,
+      "learning_rate": 4.5284430346286014e-05,
+      "loss": 1.4363,
+      "step": 17849
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6014786958694458,
+      "learning_rate": 4.5284166005815494e-05,
+      "loss": 1.3177,
+      "step": 17850
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6760414838790894,
+      "learning_rate": 4.528390166534497e-05,
+      "loss": 1.3719,
+      "step": 17851
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0424675941467285,
+      "learning_rate": 4.528363732487444e-05,
+      "loss": 1.4962,
+      "step": 17852
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9546566009521484,
+      "learning_rate": 4.5283372984403915e-05,
+      "loss": 1.4476,
+      "step": 17853
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8740450143814087,
+      "learning_rate": 4.528310864393339e-05,
+      "loss": 1.4486,
+      "step": 17854
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6727296113967896,
+      "learning_rate": 4.528284430346286e-05,
+      "loss": 1.3809,
+      "step": 17855
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.839142918586731,
+      "learning_rate": 4.5282579962992335e-05,
+      "loss": 1.4207,
+      "step": 17856
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6433759927749634,
+      "learning_rate": 4.528231562252181e-05,
+      "loss": 1.2513,
+      "step": 17857
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.983963966369629,
+      "learning_rate": 4.528205128205128e-05,
+      "loss": 1.3017,
+      "step": 17858
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.618412733078003,
+      "learning_rate": 4.5281786941580756e-05,
+      "loss": 1.3087,
+      "step": 17859
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9322333335876465,
+      "learning_rate": 4.528152260111023e-05,
+      "loss": 1.4109,
+      "step": 17860
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7666022777557373,
+      "learning_rate": 4.528125826063971e-05,
+      "loss": 1.4587,
+      "step": 17861
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.194912910461426,
+      "learning_rate": 4.528099392016918e-05,
+      "loss": 1.1973,
+      "step": 17862
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0885486602783203,
+      "learning_rate": 4.528072957969866e-05,
+      "loss": 1.3736,
+      "step": 17863
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1192660331726074,
+      "learning_rate": 4.5280465239228124e-05,
+      "loss": 1.3795,
+      "step": 17864
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5590100288391113,
+      "learning_rate": 4.5280200898757604e-05,
+      "loss": 2.444,
+      "step": 17865
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6683443784713745,
+      "learning_rate": 4.527993655828708e-05,
+      "loss": 1.3922,
+      "step": 17866
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7942850589752197,
+      "learning_rate": 4.527967221781655e-05,
+      "loss": 1.5137,
+      "step": 17867
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9481525421142578,
+      "learning_rate": 4.5279407877346025e-05,
+      "loss": 1.341,
+      "step": 17868
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.591141700744629,
+      "learning_rate": 4.52791435368755e-05,
+      "loss": 1.2226,
+      "step": 17869
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7504347562789917,
+      "learning_rate": 4.527887919640497e-05,
+      "loss": 1.2655,
+      "step": 17870
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7320809364318848,
+      "learning_rate": 4.5278614855934446e-05,
+      "loss": 1.4129,
+      "step": 17871
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7199647426605225,
+      "learning_rate": 4.527835051546392e-05,
+      "loss": 1.3534,
+      "step": 17872
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8445625305175781,
+      "learning_rate": 4.527808617499339e-05,
+      "loss": 1.4167,
+      "step": 17873
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6490862369537354,
+      "learning_rate": 4.5277821834522866e-05,
+      "loss": 1.3083,
+      "step": 17874
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6100916862487793,
+      "learning_rate": 4.527755749405234e-05,
+      "loss": 1.2145,
+      "step": 17875
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9844427108764648,
+      "learning_rate": 4.527729315358182e-05,
+      "loss": 1.935,
+      "step": 17876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4399607181549072,
+      "learning_rate": 4.527702881311129e-05,
+      "loss": 1.4259,
+      "step": 17877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.196324110031128,
+      "learning_rate": 4.527676447264077e-05,
+      "loss": 1.4468,
+      "step": 17878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7165203094482422,
+      "learning_rate": 4.5276500132170234e-05,
+      "loss": 1.4391,
+      "step": 17879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7297660112380981,
+      "learning_rate": 4.5276235791699715e-05,
+      "loss": 1.4404,
+      "step": 17880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6549034118652344,
+      "learning_rate": 4.527597145122919e-05,
+      "loss": 1.4782,
+      "step": 17881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1860854625701904,
+      "learning_rate": 4.5275707110758655e-05,
+      "loss": 1.6093,
+      "step": 17882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.799326777458191,
+      "learning_rate": 4.5275442770288135e-05,
+      "loss": 1.3226,
+      "step": 17883
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1512763500213623,
+      "learning_rate": 4.52751784298176e-05,
+      "loss": 1.562,
+      "step": 17884
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.516702651977539,
+      "learning_rate": 4.527491408934708e-05,
+      "loss": 1.6044,
+      "step": 17885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7699389457702637,
+      "learning_rate": 4.5274649748876556e-05,
+      "loss": 1.4308,
+      "step": 17886
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.856304407119751,
+      "learning_rate": 4.527438540840603e-05,
+      "loss": 1.5591,
+      "step": 17887
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5747895240783691,
+      "learning_rate": 4.52741210679355e-05,
+      "loss": 1.3204,
+      "step": 17888
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.482315182685852,
+      "learning_rate": 4.527385672746498e-05,
+      "loss": 2.2639,
+      "step": 17889
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.163360357284546,
+      "learning_rate": 4.527359238699445e-05,
+      "loss": 1.1872,
+      "step": 17890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6817165613174438,
+      "learning_rate": 4.527332804652393e-05,
+      "loss": 1.395,
+      "step": 17891
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7056593894958496,
+      "learning_rate": 4.52730637060534e-05,
+      "loss": 1.2834,
+      "step": 17892
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9342010021209717,
+      "learning_rate": 4.527279936558287e-05,
+      "loss": 1.4389,
+      "step": 17893
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0142581462860107,
+      "learning_rate": 4.5272535025112345e-05,
+      "loss": 1.3235,
+      "step": 17894
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7302358150482178,
+      "learning_rate": 4.527227068464182e-05,
+      "loss": 1.3623,
+      "step": 17895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7845104932785034,
+      "learning_rate": 4.52720063441713e-05,
+      "loss": 1.2136,
+      "step": 17896
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.025594472885132,
+      "learning_rate": 4.5271742003700765e-05,
+      "loss": 1.3295,
+      "step": 17897
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6896793842315674,
+      "learning_rate": 4.5271477663230246e-05,
+      "loss": 1.5273,
+      "step": 17898
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8527255058288574,
+      "learning_rate": 4.527121332275971e-05,
+      "loss": 1.4028,
+      "step": 17899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1901590824127197,
+      "learning_rate": 4.527094898228919e-05,
+      "loss": 1.4533,
+      "step": 17900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.782970905303955,
+      "learning_rate": 4.5270684641818666e-05,
+      "loss": 1.3596,
+      "step": 17901
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.214367151260376,
+      "learning_rate": 4.527042030134814e-05,
+      "loss": 1.3205,
+      "step": 17902
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5961087942123413,
+      "learning_rate": 4.5270155960877614e-05,
+      "loss": 1.3487,
+      "step": 17903
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9297239780426025,
+      "learning_rate": 4.526989162040709e-05,
+      "loss": 1.4829,
+      "step": 17904
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.697169303894043,
+      "learning_rate": 4.526962727993656e-05,
+      "loss": 1.3441,
+      "step": 17905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8224002122879028,
+      "learning_rate": 4.5269362939466034e-05,
+      "loss": 1.481,
+      "step": 17906
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5042861700057983,
+      "learning_rate": 4.526909859899551e-05,
+      "loss": 1.2623,
+      "step": 17907
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6918449401855469,
+      "learning_rate": 4.526883425852498e-05,
+      "loss": 1.4009,
+      "step": 17908
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.622194766998291,
+      "learning_rate": 4.5268569918054455e-05,
+      "loss": 1.2654,
+      "step": 17909
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.536120891571045,
+      "learning_rate": 4.526830557758393e-05,
+      "loss": 1.2445,
+      "step": 17910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8790004253387451,
+      "learning_rate": 4.526804123711341e-05,
+      "loss": 1.4562,
+      "step": 17911
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7050265073776245,
+      "learning_rate": 4.5267776896642876e-05,
+      "loss": 1.4579,
+      "step": 17912
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.781428575515747,
+      "learning_rate": 4.5267512556172356e-05,
+      "loss": 1.372,
+      "step": 17913
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6206175088882446,
+      "learning_rate": 4.526724821570182e-05,
+      "loss": 1.8083,
+      "step": 17914
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2401058673858643,
+      "learning_rate": 4.52669838752313e-05,
+      "loss": 1.4195,
+      "step": 17915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.776974081993103,
+      "learning_rate": 4.526671953476078e-05,
+      "loss": 1.2678,
+      "step": 17916
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9604802131652832,
+      "learning_rate": 4.526645519429025e-05,
+      "loss": 1.3229,
+      "step": 17917
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6995298862457275,
+      "learning_rate": 4.5266190853819724e-05,
+      "loss": 1.4471,
+      "step": 17918
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.840401291847229,
+      "learning_rate": 4.526592651334919e-05,
+      "loss": 1.367,
+      "step": 17919
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.086160898208618,
+      "learning_rate": 4.526566217287867e-05,
+      "loss": 1.4394,
+      "step": 17920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.657165765762329,
+      "learning_rate": 4.5265397832408145e-05,
+      "loss": 1.4965,
+      "step": 17921
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7606759071350098,
+      "learning_rate": 4.526513349193762e-05,
+      "loss": 1.4178,
+      "step": 17922
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7641103267669678,
+      "learning_rate": 4.526486915146709e-05,
+      "loss": 1.1758,
+      "step": 17923
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7915714979171753,
+      "learning_rate": 4.5264604810996565e-05,
+      "loss": 1.2394,
+      "step": 17924
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.138134479522705,
+      "learning_rate": 4.526434047052604e-05,
+      "loss": 1.5725,
+      "step": 17925
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3677759170532227,
+      "learning_rate": 4.526407613005552e-05,
+      "loss": 1.4146,
+      "step": 17926
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7877956628799438,
+      "learning_rate": 4.5263811789584986e-05,
+      "loss": 1.3228,
+      "step": 17927
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.752026915550232,
+      "learning_rate": 4.526354744911446e-05,
+      "loss": 1.4269,
+      "step": 17928
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7104580402374268,
+      "learning_rate": 4.526328310864393e-05,
+      "loss": 1.3166,
+      "step": 17929
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2474029064178467,
+      "learning_rate": 4.526301876817341e-05,
+      "loss": 1.2112,
+      "step": 17930
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.044429302215576,
+      "learning_rate": 4.526275442770289e-05,
+      "loss": 1.2094,
+      "step": 17931
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6004339456558228,
+      "learning_rate": 4.5262490087232354e-05,
+      "loss": 1.4058,
+      "step": 17932
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1902730464935303,
+      "learning_rate": 4.5262225746761834e-05,
+      "loss": 1.5431,
+      "step": 17933
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8714452981948853,
+      "learning_rate": 4.52619614062913e-05,
+      "loss": 1.4779,
+      "step": 17934
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7756742238998413,
+      "learning_rate": 4.526169706582078e-05,
+      "loss": 1.3513,
+      "step": 17935
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6553668975830078,
+      "learning_rate": 4.5261432725350255e-05,
+      "loss": 1.4115,
+      "step": 17936
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6273213624954224,
+      "learning_rate": 4.526116838487973e-05,
+      "loss": 1.2683,
+      "step": 17937
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.056256055831909,
+      "learning_rate": 4.52609040444092e-05,
+      "loss": 1.404,
+      "step": 17938
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9415574073791504,
+      "learning_rate": 4.5260639703938676e-05,
+      "loss": 1.488,
+      "step": 17939
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.851062297821045,
+      "learning_rate": 4.526037536346815e-05,
+      "loss": 1.3069,
+      "step": 17940
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.753920316696167,
+      "learning_rate": 4.526011102299762e-05,
+      "loss": 1.5438,
+      "step": 17941
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8156794309616089,
+      "learning_rate": 4.5259846682527096e-05,
+      "loss": 1.5163,
+      "step": 17942
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0772697925567627,
+      "learning_rate": 4.525958234205657e-05,
+      "loss": 1.5083,
+      "step": 17943
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7182762622833252,
+      "learning_rate": 4.5259318001586044e-05,
+      "loss": 1.3224,
+      "step": 17944
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8097947835922241,
+      "learning_rate": 4.525905366111552e-05,
+      "loss": 1.2564,
+      "step": 17945
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6393609046936035,
+      "learning_rate": 4.5258789320645e-05,
+      "loss": 1.1902,
+      "step": 17946
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6132466793060303,
+      "learning_rate": 4.5258524980174464e-05,
+      "loss": 1.541,
+      "step": 17947
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7233257293701172,
+      "learning_rate": 4.5258260639703945e-05,
+      "loss": 1.2548,
+      "step": 17948
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4923348426818848,
+      "learning_rate": 4.525799629923341e-05,
+      "loss": 1.3483,
+      "step": 17949
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.423189640045166,
+      "learning_rate": 4.525773195876289e-05,
+      "loss": 2.1774,
+      "step": 17950
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.664984941482544,
+      "learning_rate": 4.5257467618292365e-05,
+      "loss": 1.416,
+      "step": 17951
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8197274208068848,
+      "learning_rate": 4.525720327782184e-05,
+      "loss": 1.4833,
+      "step": 17952
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0618410110473633,
+      "learning_rate": 4.525693893735131e-05,
+      "loss": 1.5571,
+      "step": 17953
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9452049732208252,
+      "learning_rate": 4.525667459688078e-05,
+      "loss": 1.2892,
+      "step": 17954
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7132643461227417,
+      "learning_rate": 4.525641025641026e-05,
+      "loss": 1.4185,
+      "step": 17955
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7477744817733765,
+      "learning_rate": 4.525614591593973e-05,
+      "loss": 1.3027,
+      "step": 17956
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2717151641845703,
+      "learning_rate": 4.525588157546921e-05,
+      "loss": 1.4665,
+      "step": 17957
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3851977586746216,
+      "learning_rate": 4.525561723499868e-05,
+      "loss": 1.3812,
+      "step": 17958
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7151745557785034,
+      "learning_rate": 4.5255352894528154e-05,
+      "loss": 1.4382,
+      "step": 17959
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.707431674003601,
+      "learning_rate": 4.525508855405763e-05,
+      "loss": 1.2296,
+      "step": 17960
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6430222988128662,
+      "learning_rate": 4.525482421358711e-05,
+      "loss": 2.0141,
+      "step": 17961
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9117566347122192,
+      "learning_rate": 4.5254559873116575e-05,
+      "loss": 1.356,
+      "step": 17962
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6429158449172974,
+      "learning_rate": 4.5254295532646055e-05,
+      "loss": 1.2827,
+      "step": 17963
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8174171447753906,
+      "learning_rate": 4.525403119217552e-05,
+      "loss": 1.3366,
+      "step": 17964
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7999480962753296,
+      "learning_rate": 4.5253766851704995e-05,
+      "loss": 1.3328,
+      "step": 17965
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8474925756454468,
+      "learning_rate": 4.5253502511234476e-05,
+      "loss": 1.354,
+      "step": 17966
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5711901187896729,
+      "learning_rate": 4.525323817076394e-05,
+      "loss": 1.4304,
+      "step": 17967
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.092600107192993,
+      "learning_rate": 4.525297383029342e-05,
+      "loss": 1.5366,
+      "step": 17968
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6950056552886963,
+      "learning_rate": 4.525270948982289e-05,
+      "loss": 1.404,
+      "step": 17969
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3409086465835571,
+      "learning_rate": 4.525244514935237e-05,
+      "loss": 1.347,
+      "step": 17970
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9593201875686646,
+      "learning_rate": 4.5252180808881843e-05,
+      "loss": 1.322,
+      "step": 17971
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.787627100944519,
+      "learning_rate": 4.525191646841132e-05,
+      "loss": 1.3519,
+      "step": 17972
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.961869239807129,
+      "learning_rate": 4.525165212794079e-05,
+      "loss": 1.2751,
+      "step": 17973
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7621325254440308,
+      "learning_rate": 4.5251387787470264e-05,
+      "loss": 1.3358,
+      "step": 17974
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7900325059890747,
+      "learning_rate": 4.525112344699974e-05,
+      "loss": 2.6861,
+      "step": 17975
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5674197673797607,
+      "learning_rate": 4.525085910652921e-05,
+      "loss": 1.8929,
+      "step": 17976
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1448326110839844,
+      "learning_rate": 4.5250594766058685e-05,
+      "loss": 1.2978,
+      "step": 17977
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.168994903564453,
+      "learning_rate": 4.525033042558816e-05,
+      "loss": 1.5751,
+      "step": 17978
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9562336206436157,
+      "learning_rate": 4.525006608511763e-05,
+      "loss": 1.3467,
+      "step": 17979
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4523180723190308,
+      "learning_rate": 4.5249801744647106e-05,
+      "loss": 1.2993,
+      "step": 17980
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4132180213928223,
+      "learning_rate": 4.5249537404176586e-05,
+      "loss": 1.3216,
+      "step": 17981
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8038350343704224,
+      "learning_rate": 4.524927306370605e-05,
+      "loss": 1.36,
+      "step": 17982
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5981390476226807,
+      "learning_rate": 4.524900872323553e-05,
+      "loss": 1.3473,
+      "step": 17983
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7757840156555176,
+      "learning_rate": 4.5248744382765e-05,
+      "loss": 2.0684,
+      "step": 17984
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.000408411026001,
+      "learning_rate": 4.524848004229448e-05,
+      "loss": 1.4464,
+      "step": 17985
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7396676540374756,
+      "learning_rate": 4.5248215701823954e-05,
+      "loss": 1.361,
+      "step": 17986
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6340659856796265,
+      "learning_rate": 4.524795136135343e-05,
+      "loss": 1.3159,
+      "step": 17987
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.648041248321533,
+      "learning_rate": 4.52476870208829e-05,
+      "loss": 1.4534,
+      "step": 17988
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3424177169799805,
+      "learning_rate": 4.5247422680412375e-05,
+      "loss": 1.6995,
+      "step": 17989
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9552439451217651,
+      "learning_rate": 4.524715833994185e-05,
+      "loss": 1.3684,
+      "step": 17990
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6948133707046509,
+      "learning_rate": 4.524689399947132e-05,
+      "loss": 1.5039,
+      "step": 17991
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5975024700164795,
+      "learning_rate": 4.5246629659000795e-05,
+      "loss": 1.3756,
+      "step": 17992
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8171504735946655,
+      "learning_rate": 4.524636531853027e-05,
+      "loss": 1.6015,
+      "step": 17993
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9162218570709229,
+      "learning_rate": 4.524610097805974e-05,
+      "loss": 1.3637,
+      "step": 17994
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9026254415512085,
+      "learning_rate": 4.5245836637589216e-05,
+      "loss": 1.4016,
+      "step": 17995
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5983543395996094,
+      "learning_rate": 4.5245572297118696e-05,
+      "loss": 1.5554,
+      "step": 17996
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6374788284301758,
+      "learning_rate": 4.524530795664816e-05,
+      "loss": 1.4486,
+      "step": 17997
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.688722014427185,
+      "learning_rate": 4.5245043616177643e-05,
+      "loss": 1.4019,
+      "step": 17998
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1011719703674316,
+      "learning_rate": 4.524477927570711e-05,
+      "loss": 1.3672,
+      "step": 17999
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4581266641616821,
+      "learning_rate": 4.5244514935236584e-05,
+      "loss": 1.2384,
+      "step": 18000
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8911775350570679,
+      "learning_rate": 4.5244250594766064e-05,
+      "loss": 1.1803,
+      "step": 18001
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0361087322235107,
+      "learning_rate": 4.524398625429553e-05,
+      "loss": 1.3382,
+      "step": 18002
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8335589170455933,
+      "learning_rate": 4.524372191382501e-05,
+      "loss": 1.5005,
+      "step": 18003
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0778775215148926,
+      "learning_rate": 4.524345757335448e-05,
+      "loss": 1.3163,
+      "step": 18004
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.950873613357544,
+      "learning_rate": 4.524319323288396e-05,
+      "loss": 1.4594,
+      "step": 18005
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7673248052597046,
+      "learning_rate": 4.524292889241343e-05,
+      "loss": 1.3359,
+      "step": 18006
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6008762121200562,
+      "learning_rate": 4.5242664551942906e-05,
+      "loss": 1.3564,
+      "step": 18007
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.875124454498291,
+      "learning_rate": 4.524240021147238e-05,
+      "loss": 1.381,
+      "step": 18008
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1520791053771973,
+      "learning_rate": 4.524213587100185e-05,
+      "loss": 1.3631,
+      "step": 18009
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9877640008926392,
+      "learning_rate": 4.5241871530531326e-05,
+      "loss": 1.4597,
+      "step": 18010
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1388821601867676,
+      "learning_rate": 4.52416071900608e-05,
+      "loss": 1.5478,
+      "step": 18011
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.494380235671997,
+      "learning_rate": 4.5241342849590273e-05,
+      "loss": 1.3231,
+      "step": 18012
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.45628821849823,
+      "learning_rate": 4.524107850911975e-05,
+      "loss": 1.3301,
+      "step": 18013
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6632179021835327,
+      "learning_rate": 4.524081416864922e-05,
+      "loss": 1.4075,
+      "step": 18014
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3205621242523193,
+      "learning_rate": 4.5240549828178694e-05,
+      "loss": 1.5069,
+      "step": 18015
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1579320430755615,
+      "learning_rate": 4.5240285487708175e-05,
+      "loss": 1.4887,
+      "step": 18016
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0371944904327393,
+      "learning_rate": 4.524002114723764e-05,
+      "loss": 1.4862,
+      "step": 18017
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8026766777038574,
+      "learning_rate": 4.523975680676712e-05,
+      "loss": 1.5162,
+      "step": 18018
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.800262451171875,
+      "learning_rate": 4.523949246629659e-05,
+      "loss": 1.4091,
+      "step": 18019
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0222601890563965,
+      "learning_rate": 4.523922812582607e-05,
+      "loss": 1.3513,
+      "step": 18020
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0040807723999023,
+      "learning_rate": 4.523896378535554e-05,
+      "loss": 1.5025,
+      "step": 18021
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1030590534210205,
+      "learning_rate": 4.5238699444885016e-05,
+      "loss": 1.3008,
+      "step": 18022
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6869451999664307,
+      "learning_rate": 4.523843510441449e-05,
+      "loss": 1.3714,
+      "step": 18023
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.021068811416626,
+      "learning_rate": 4.523817076394396e-05,
+      "loss": 1.4905,
+      "step": 18024
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5948525667190552,
+      "learning_rate": 4.523790642347344e-05,
+      "loss": 1.4108,
+      "step": 18025
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7416374683380127,
+      "learning_rate": 4.523764208300291e-05,
+      "loss": 2.778,
+      "step": 18026
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.052192211151123,
+      "learning_rate": 4.5237377742532384e-05,
+      "loss": 1.376,
+      "step": 18027
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5798215866088867,
+      "learning_rate": 4.523711340206186e-05,
+      "loss": 2.1217,
+      "step": 18028
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.195298671722412,
+      "learning_rate": 4.523684906159133e-05,
+      "loss": 1.3456,
+      "step": 18029
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6918030977249146,
+      "learning_rate": 4.5236584721120804e-05,
+      "loss": 1.4507,
+      "step": 18030
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5934910774230957,
+      "learning_rate": 4.5236320380650285e-05,
+      "loss": 2.4024,
+      "step": 18031
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5256403684616089,
+      "learning_rate": 4.523605604017975e-05,
+      "loss": 1.2856,
+      "step": 18032
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1970627307891846,
+      "learning_rate": 4.523579169970923e-05,
+      "loss": 1.4515,
+      "step": 18033
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.334977388381958,
+      "learning_rate": 4.52355273592387e-05,
+      "loss": 1.2693,
+      "step": 18034
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0685105323791504,
+      "learning_rate": 4.523526301876818e-05,
+      "loss": 1.6232,
+      "step": 18035
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6652997732162476,
+      "learning_rate": 4.523499867829765e-05,
+      "loss": 1.3684,
+      "step": 18036
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9364243745803833,
+      "learning_rate": 4.523473433782712e-05,
+      "loss": 1.5437,
+      "step": 18037
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7321858406066895,
+      "learning_rate": 4.52344699973566e-05,
+      "loss": 1.3999,
+      "step": 18038
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3695569038391113,
+      "learning_rate": 4.523420565688607e-05,
+      "loss": 1.371,
+      "step": 18039
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4416120052337646,
+      "learning_rate": 4.523394131641555e-05,
+      "loss": 2.1264,
+      "step": 18040
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2637939453125,
+      "learning_rate": 4.523367697594502e-05,
+      "loss": 1.4234,
+      "step": 18041
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.216158866882324,
+      "learning_rate": 4.5233412635474494e-05,
+      "loss": 1.4866,
+      "step": 18042
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4927055835723877,
+      "learning_rate": 4.523314829500397e-05,
+      "loss": 1.527,
+      "step": 18043
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5315659046173096,
+      "learning_rate": 4.523288395453344e-05,
+      "loss": 1.3078,
+      "step": 18044
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7329617738723755,
+      "learning_rate": 4.5232619614062915e-05,
+      "loss": 1.4829,
+      "step": 18045
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8757516145706177,
+      "learning_rate": 4.523235527359239e-05,
+      "loss": 1.4891,
+      "step": 18046
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9763453006744385,
+      "learning_rate": 4.523209093312186e-05,
+      "loss": 1.4122,
+      "step": 18047
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6679478883743286,
+      "learning_rate": 4.5231826592651336e-05,
+      "loss": 1.405,
+      "step": 18048
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.892820119857788,
+      "learning_rate": 4.523156225218081e-05,
+      "loss": 1.366,
+      "step": 18049
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5959174633026123,
+      "learning_rate": 4.523129791171028e-05,
+      "loss": 1.2977,
+      "step": 18050
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9270761013031006,
+      "learning_rate": 4.523103357123976e-05,
+      "loss": 1.2074,
+      "step": 18051
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0263164043426514,
+      "learning_rate": 4.523076923076923e-05,
+      "loss": 1.5282,
+      "step": 18052
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8126249313354492,
+      "learning_rate": 4.523050489029871e-05,
+      "loss": 1.4973,
+      "step": 18053
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.969761848449707,
+      "learning_rate": 4.523024054982818e-05,
+      "loss": 1.4453,
+      "step": 18054
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0495293140411377,
+      "learning_rate": 4.522997620935766e-05,
+      "loss": 1.3981,
+      "step": 18055
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.525894284248352,
+      "learning_rate": 4.522971186888713e-05,
+      "loss": 1.3636,
+      "step": 18056
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6473486423492432,
+      "learning_rate": 4.5229447528416604e-05,
+      "loss": 1.258,
+      "step": 18057
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7892813682556152,
+      "learning_rate": 4.522918318794608e-05,
+      "loss": 1.3857,
+      "step": 18058
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6621049642562866,
+      "learning_rate": 4.522891884747555e-05,
+      "loss": 1.4531,
+      "step": 18059
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7124232053756714,
+      "learning_rate": 4.5228654507005025e-05,
+      "loss": 1.228,
+      "step": 18060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.637648582458496,
+      "learning_rate": 4.52283901665345e-05,
+      "loss": 1.2534,
+      "step": 18061
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8033677339553833,
+      "learning_rate": 4.522812582606397e-05,
+      "loss": 1.5067,
+      "step": 18062
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.447164297103882,
+      "learning_rate": 4.5227861485593446e-05,
+      "loss": 1.3728,
+      "step": 18063
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5619441270828247,
+      "learning_rate": 4.522759714512292e-05,
+      "loss": 1.2899,
+      "step": 18064
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9613438844680786,
+      "learning_rate": 4.522733280465239e-05,
+      "loss": 1.4622,
+      "step": 18065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9303041696548462,
+      "learning_rate": 4.522706846418187e-05,
+      "loss": 1.2361,
+      "step": 18066
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.365606427192688,
+      "learning_rate": 4.522680412371134e-05,
+      "loss": 1.3779,
+      "step": 18067
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6397703886032104,
+      "learning_rate": 4.522653978324082e-05,
+      "loss": 1.3472,
+      "step": 18068
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.79519784450531,
+      "learning_rate": 4.522627544277029e-05,
+      "loss": 1.3943,
+      "step": 18069
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7750693559646606,
+      "learning_rate": 4.522601110229977e-05,
+      "loss": 1.3502,
+      "step": 18070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4335150718688965,
+      "learning_rate": 4.522574676182924e-05,
+      "loss": 1.2412,
+      "step": 18071
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4889761209487915,
+      "learning_rate": 4.522548242135871e-05,
+      "loss": 1.3818,
+      "step": 18072
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3933234214782715,
+      "learning_rate": 4.522521808088819e-05,
+      "loss": 1.2097,
+      "step": 18073
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.582700729370117,
+      "learning_rate": 4.5224953740417655e-05,
+      "loss": 1.3284,
+      "step": 18074
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.72604238986969,
+      "learning_rate": 4.5224689399947136e-05,
+      "loss": 1.3609,
+      "step": 18075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6492893695831299,
+      "learning_rate": 4.522442505947661e-05,
+      "loss": 1.3715,
+      "step": 18076
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0335025787353516,
+      "learning_rate": 4.522416071900608e-05,
+      "loss": 1.5809,
+      "step": 18077
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.02958083152771,
+      "learning_rate": 4.5223896378535556e-05,
+      "loss": 1.8483,
+      "step": 18078
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4798918962478638,
+      "learning_rate": 4.522363203806503e-05,
+      "loss": 1.3853,
+      "step": 18079
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6028368473052979,
+      "learning_rate": 4.52233676975945e-05,
+      "loss": 1.5288,
+      "step": 18080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4934076070785522,
+      "learning_rate": 4.5223103357123984e-05,
+      "loss": 1.9815,
+      "step": 18081
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.455379605293274,
+      "learning_rate": 4.522283901665345e-05,
+      "loss": 1.2184,
+      "step": 18082
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4275598526000977,
+      "learning_rate": 4.5222574676182924e-05,
+      "loss": 1.287,
+      "step": 18083
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4814345836639404,
+      "learning_rate": 4.52223103357124e-05,
+      "loss": 1.2776,
+      "step": 18084
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8318698406219482,
+      "learning_rate": 4.522204599524187e-05,
+      "loss": 1.3323,
+      "step": 18085
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9421546459197998,
+      "learning_rate": 4.522178165477135e-05,
+      "loss": 1.2772,
+      "step": 18086
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1705801486968994,
+      "learning_rate": 4.522151731430082e-05,
+      "loss": 1.3636,
+      "step": 18087
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7980334758758545,
+      "learning_rate": 4.52212529738303e-05,
+      "loss": 1.2998,
+      "step": 18088
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9345948696136475,
+      "learning_rate": 4.5220988633359766e-05,
+      "loss": 1.5527,
+      "step": 18089
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9105020761489868,
+      "learning_rate": 4.5220724292889246e-05,
+      "loss": 1.4645,
+      "step": 18090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5586451292037964,
+      "learning_rate": 4.522045995241872e-05,
+      "loss": 1.334,
+      "step": 18091
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9253586530685425,
+      "learning_rate": 4.522019561194819e-05,
+      "loss": 1.3455,
+      "step": 18092
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8313612937927246,
+      "learning_rate": 4.5219931271477667e-05,
+      "loss": 1.7107,
+      "step": 18093
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8371528387069702,
+      "learning_rate": 4.521966693100714e-05,
+      "loss": 1.5013,
+      "step": 18094
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0543580055236816,
+      "learning_rate": 4.5219402590536614e-05,
+      "loss": 1.461,
+      "step": 18095
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8440587520599365,
+      "learning_rate": 4.521913825006609e-05,
+      "loss": 1.4375,
+      "step": 18096
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3310556411743164,
+      "learning_rate": 4.521887390959556e-05,
+      "loss": 1.4155,
+      "step": 18097
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.371201992034912,
+      "learning_rate": 4.5218609569125034e-05,
+      "loss": 1.3513,
+      "step": 18098
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4481984376907349,
+      "learning_rate": 4.521834522865451e-05,
+      "loss": 1.3857,
+      "step": 18099
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6244150400161743,
+      "learning_rate": 4.521808088818398e-05,
+      "loss": 1.4551,
+      "step": 18100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3317389488220215,
+      "learning_rate": 4.521781654771346e-05,
+      "loss": 1.5432,
+      "step": 18101
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7836248874664307,
+      "learning_rate": 4.521755220724293e-05,
+      "loss": 1.3924,
+      "step": 18102
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9559710025787354,
+      "learning_rate": 4.521728786677241e-05,
+      "loss": 1.397,
+      "step": 18103
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.533209204673767,
+      "learning_rate": 4.5217023526301876e-05,
+      "loss": 1.3564,
+      "step": 18104
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6495758295059204,
+      "learning_rate": 4.5216759185831356e-05,
+      "loss": 2.0962,
+      "step": 18105
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.721641182899475,
+      "learning_rate": 4.521649484536083e-05,
+      "loss": 1.4031,
+      "step": 18106
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6819837093353271,
+      "learning_rate": 4.52162305048903e-05,
+      "loss": 1.2836,
+      "step": 18107
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9419044256210327,
+      "learning_rate": 4.521596616441978e-05,
+      "loss": 1.4496,
+      "step": 18108
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3923274278640747,
+      "learning_rate": 4.5215701823949244e-05,
+      "loss": 1.261,
+      "step": 18109
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5816811323165894,
+      "learning_rate": 4.5215437483478724e-05,
+      "loss": 1.3933,
+      "step": 18110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.840917706489563,
+      "learning_rate": 4.52151731430082e-05,
+      "loss": 1.3041,
+      "step": 18111
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5944174528121948,
+      "learning_rate": 4.521490880253767e-05,
+      "loss": 1.4297,
+      "step": 18112
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5985281467437744,
+      "learning_rate": 4.5214644462067145e-05,
+      "loss": 1.5949,
+      "step": 18113
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0056769847869873,
+      "learning_rate": 4.521438012159662e-05,
+      "loss": 1.4148,
+      "step": 18114
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0998780727386475,
+      "learning_rate": 4.521411578112609e-05,
+      "loss": 1.3608,
+      "step": 18115
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7739454507827759,
+      "learning_rate": 4.521385144065557e-05,
+      "loss": 1.9263,
+      "step": 18116
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.602275013923645,
+      "learning_rate": 4.521358710018504e-05,
+      "loss": 1.3601,
+      "step": 18117
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5970892906188965,
+      "learning_rate": 4.521332275971451e-05,
+      "loss": 1.3034,
+      "step": 18118
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.193080425262451,
+      "learning_rate": 4.5213058419243986e-05,
+      "loss": 1.4735,
+      "step": 18119
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8842791318893433,
+      "learning_rate": 4.521279407877346e-05,
+      "loss": 1.2305,
+      "step": 18120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.464198112487793,
+      "learning_rate": 4.521252973830294e-05,
+      "loss": 2.1601,
+      "step": 18121
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7098255157470703,
+      "learning_rate": 4.521226539783241e-05,
+      "loss": 1.3218,
+      "step": 18122
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8503955602645874,
+      "learning_rate": 4.521200105736189e-05,
+      "loss": 1.4979,
+      "step": 18123
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.977622389793396,
+      "learning_rate": 4.5211736716891354e-05,
+      "loss": 1.5567,
+      "step": 18124
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6812856197357178,
+      "learning_rate": 4.5211472376420834e-05,
+      "loss": 1.4038,
+      "step": 18125
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.159296751022339,
+      "learning_rate": 4.521120803595031e-05,
+      "loss": 1.5075,
+      "step": 18126
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.026224374771118,
+      "learning_rate": 4.521094369547978e-05,
+      "loss": 1.4869,
+      "step": 18127
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7739681005477905,
+      "learning_rate": 4.5210679355009255e-05,
+      "loss": 1.357,
+      "step": 18128
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.749954342842102,
+      "learning_rate": 4.521041501453873e-05,
+      "loss": 1.3199,
+      "step": 18129
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.554189920425415,
+      "learning_rate": 4.52101506740682e-05,
+      "loss": 1.3405,
+      "step": 18130
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6914371252059937,
+      "learning_rate": 4.5209886333597676e-05,
+      "loss": 1.3621,
+      "step": 18131
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5982463359832764,
+      "learning_rate": 4.520962199312715e-05,
+      "loss": 2.1335,
+      "step": 18132
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3205976486206055,
+      "learning_rate": 4.520935765265662e-05,
+      "loss": 1.6405,
+      "step": 18133
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8487590551376343,
+      "learning_rate": 4.5209093312186097e-05,
+      "loss": 1.3517,
+      "step": 18134
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8948386907577515,
+      "learning_rate": 4.520882897171557e-05,
+      "loss": 1.279,
+      "step": 18135
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5091805458068848,
+      "learning_rate": 4.520856463124505e-05,
+      "loss": 2.3237,
+      "step": 18136
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8103984594345093,
+      "learning_rate": 4.520830029077452e-05,
+      "loss": 1.4282,
+      "step": 18137
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3066729307174683,
+      "learning_rate": 4.5208035950304e-05,
+      "loss": 1.2055,
+      "step": 18138
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4094655513763428,
+      "learning_rate": 4.5207771609833464e-05,
+      "loss": 1.1359,
+      "step": 18139
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9255043268203735,
+      "learning_rate": 4.5207507269362945e-05,
+      "loss": 1.414,
+      "step": 18140
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3602898120880127,
+      "learning_rate": 4.520724292889242e-05,
+      "loss": 1.4599,
+      "step": 18141
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1989686489105225,
+      "learning_rate": 4.520697858842189e-05,
+      "loss": 1.4561,
+      "step": 18142
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.851233720779419,
+      "learning_rate": 4.5206714247951365e-05,
+      "loss": 1.3651,
+      "step": 18143
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7548974752426147,
+      "learning_rate": 4.520644990748084e-05,
+      "loss": 1.3643,
+      "step": 18144
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.038209915161133,
+      "learning_rate": 4.520618556701031e-05,
+      "loss": 1.4573,
+      "step": 18145
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3564627170562744,
+      "learning_rate": 4.5205921226539786e-05,
+      "loss": 1.6435,
+      "step": 18146
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.113607883453369,
+      "learning_rate": 4.520565688606926e-05,
+      "loss": 1.388,
+      "step": 18147
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.582930088043213,
+      "learning_rate": 4.520539254559873e-05,
+      "loss": 1.3194,
+      "step": 18148
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8635331392288208,
+      "learning_rate": 4.520512820512821e-05,
+      "loss": 1.27,
+      "step": 18149
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0916855335235596,
+      "learning_rate": 4.520486386465768e-05,
+      "loss": 1.3981,
+      "step": 18150
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9590967893600464,
+      "learning_rate": 4.520459952418716e-05,
+      "loss": 1.4584,
+      "step": 18151
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6896672248840332,
+      "learning_rate": 4.520433518371663e-05,
+      "loss": 1.3968,
+      "step": 18152
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6241470575332642,
+      "learning_rate": 4.520407084324611e-05,
+      "loss": 1.3385,
+      "step": 18153
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1432292461395264,
+      "learning_rate": 4.5203806502775575e-05,
+      "loss": 1.4686,
+      "step": 18154
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0607376098632812,
+      "learning_rate": 4.520354216230505e-05,
+      "loss": 1.4375,
+      "step": 18155
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.806001901626587,
+      "learning_rate": 4.520327782183453e-05,
+      "loss": 1.5557,
+      "step": 18156
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9998613595962524,
+      "learning_rate": 4.5203013481363995e-05,
+      "loss": 1.4036,
+      "step": 18157
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9124106168746948,
+      "learning_rate": 4.5202749140893476e-05,
+      "loss": 1.4882,
+      "step": 18158
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0678651332855225,
+      "learning_rate": 4.520248480042294e-05,
+      "loss": 1.4566,
+      "step": 18159
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7426021099090576,
+      "learning_rate": 4.520222045995242e-05,
+      "loss": 1.6873,
+      "step": 18160
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7468756437301636,
+      "learning_rate": 4.5201956119481896e-05,
+      "loss": 1.3222,
+      "step": 18161
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7180043458938599,
+      "learning_rate": 4.520169177901137e-05,
+      "loss": 1.2251,
+      "step": 18162
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.892011046409607,
+      "learning_rate": 4.5201427438540844e-05,
+      "loss": 1.5327,
+      "step": 18163
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6189329624176025,
+      "learning_rate": 4.520116309807032e-05,
+      "loss": 1.2326,
+      "step": 18164
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.407594919204712,
+      "learning_rate": 4.520089875759979e-05,
+      "loss": 1.359,
+      "step": 18165
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3658194541931152,
+      "learning_rate": 4.5200634417129264e-05,
+      "loss": 1.5353,
+      "step": 18166
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5371543169021606,
+      "learning_rate": 4.520037007665874e-05,
+      "loss": 1.2416,
+      "step": 18167
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1569318771362305,
+      "learning_rate": 4.520010573618821e-05,
+      "loss": 1.4175,
+      "step": 18168
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5898040533065796,
+      "learning_rate": 4.5199841395717685e-05,
+      "loss": 1.2379,
+      "step": 18169
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8996071815490723,
+      "learning_rate": 4.519957705524716e-05,
+      "loss": 1.4509,
+      "step": 18170
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.331019878387451,
+      "learning_rate": 4.519931271477664e-05,
+      "loss": 1.48,
+      "step": 18171
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5882669687271118,
+      "learning_rate": 4.5199048374306106e-05,
+      "loss": 1.2196,
+      "step": 18172
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3091964721679688,
+      "learning_rate": 4.5198784033835586e-05,
+      "loss": 1.2553,
+      "step": 18173
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5008667707443237,
+      "learning_rate": 4.519851969336505e-05,
+      "loss": 2.5058,
+      "step": 18174
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6233731508255005,
+      "learning_rate": 4.519825535289453e-05,
+      "loss": 1.3738,
+      "step": 18175
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6828579902648926,
+      "learning_rate": 4.519799101242401e-05,
+      "loss": 1.3447,
+      "step": 18176
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8622713088989258,
+      "learning_rate": 4.519772667195348e-05,
+      "loss": 1.6648,
+      "step": 18177
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.037503480911255,
+      "learning_rate": 4.5197462331482954e-05,
+      "loss": 1.5133,
+      "step": 18178
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6780949831008911,
+      "learning_rate": 4.519719799101243e-05,
+      "loss": 1.4176,
+      "step": 18179
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6760156154632568,
+      "learning_rate": 4.51969336505419e-05,
+      "loss": 1.296,
+      "step": 18180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9891365766525269,
+      "learning_rate": 4.5196669310071375e-05,
+      "loss": 1.4855,
+      "step": 18181
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.026427745819092,
+      "learning_rate": 4.519640496960085e-05,
+      "loss": 1.5609,
+      "step": 18182
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.125182628631592,
+      "learning_rate": 4.519614062913032e-05,
+      "loss": 1.4167,
+      "step": 18183
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9404304027557373,
+      "learning_rate": 4.5195876288659795e-05,
+      "loss": 1.4643,
+      "step": 18184
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8017899990081787,
+      "learning_rate": 4.519561194818927e-05,
+      "loss": 1.4485,
+      "step": 18185
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7958874702453613,
+      "learning_rate": 4.519534760771875e-05,
+      "loss": 1.5296,
+      "step": 18186
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9908676147460938,
+      "learning_rate": 4.5195083267248216e-05,
+      "loss": 1.2738,
+      "step": 18187
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5525487661361694,
+      "learning_rate": 4.5194818926777696e-05,
+      "loss": 1.4052,
+      "step": 18188
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6491600275039673,
+      "learning_rate": 4.519455458630716e-05,
+      "loss": 1.1883,
+      "step": 18189
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.053197145462036,
+      "learning_rate": 4.519429024583664e-05,
+      "loss": 2.4141,
+      "step": 18190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9291672706604004,
+      "learning_rate": 4.519402590536612e-05,
+      "loss": 1.6212,
+      "step": 18191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5565698146820068,
+      "learning_rate": 4.5193761564895584e-05,
+      "loss": 1.4471,
+      "step": 18192
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6567609310150146,
+      "learning_rate": 4.5193497224425064e-05,
+      "loss": 1.3005,
+      "step": 18193
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6558799743652344,
+      "learning_rate": 4.519323288395453e-05,
+      "loss": 1.4293,
+      "step": 18194
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.780643343925476,
+      "learning_rate": 4.519296854348401e-05,
+      "loss": 1.4373,
+      "step": 18195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.846806526184082,
+      "learning_rate": 4.5192704203013485e-05,
+      "loss": 1.3914,
+      "step": 18196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5137982368469238,
+      "learning_rate": 4.519243986254296e-05,
+      "loss": 1.3838,
+      "step": 18197
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.267411231994629,
+      "learning_rate": 4.519217552207243e-05,
+      "loss": 1.6328,
+      "step": 18198
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2514376640319824,
+      "learning_rate": 4.5191911181601906e-05,
+      "loss": 1.5361,
+      "step": 18199
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2815847396850586,
+      "learning_rate": 4.519164684113138e-05,
+      "loss": 1.4623,
+      "step": 18200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.819046139717102,
+      "learning_rate": 4.519138250066085e-05,
+      "loss": 1.3914,
+      "step": 18201
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.151395797729492,
+      "learning_rate": 4.5191118160190326e-05,
+      "loss": 1.434,
+      "step": 18202
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5558555126190186,
+      "learning_rate": 4.51908538197198e-05,
+      "loss": 1.2877,
+      "step": 18203
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.988766074180603,
+      "learning_rate": 4.5190589479249274e-05,
+      "loss": 1.3047,
+      "step": 18204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9633381366729736,
+      "learning_rate": 4.519032513877875e-05,
+      "loss": 1.3158,
+      "step": 18205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7961493730545044,
+      "learning_rate": 4.519006079830823e-05,
+      "loss": 1.3775,
+      "step": 18206
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.045325994491577,
+      "learning_rate": 4.5189796457837694e-05,
+      "loss": 1.5068,
+      "step": 18207
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8279575109481812,
+      "learning_rate": 4.5189532117367175e-05,
+      "loss": 1.4311,
+      "step": 18208
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3906772136688232,
+      "learning_rate": 4.518926777689664e-05,
+      "loss": 1.5065,
+      "step": 18209
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9218320846557617,
+      "learning_rate": 4.518900343642612e-05,
+      "loss": 1.4035,
+      "step": 18210
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6401885747909546,
+      "learning_rate": 4.5188739095955595e-05,
+      "loss": 1.4272,
+      "step": 18211
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5517363548278809,
+      "learning_rate": 4.518847475548507e-05,
+      "loss": 1.3527,
+      "step": 18212
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8545951843261719,
+      "learning_rate": 4.518821041501454e-05,
+      "loss": 1.1713,
+      "step": 18213
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7204912900924683,
+      "learning_rate": 4.5187946074544016e-05,
+      "loss": 1.3695,
+      "step": 18214
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3886938095092773,
+      "learning_rate": 4.518768173407349e-05,
+      "loss": 1.4204,
+      "step": 18215
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.115269422531128,
+      "learning_rate": 4.518741739360296e-05,
+      "loss": 1.2785,
+      "step": 18216
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7003538608551025,
+      "learning_rate": 4.518715305313244e-05,
+      "loss": 1.4107,
+      "step": 18217
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5612164735794067,
+      "learning_rate": 4.518688871266191e-05,
+      "loss": 1.4045,
+      "step": 18218
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3300458192825317,
+      "learning_rate": 4.5186624372191384e-05,
+      "loss": 1.3685,
+      "step": 18219
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5559452772140503,
+      "learning_rate": 4.518636003172086e-05,
+      "loss": 1.3775,
+      "step": 18220
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.61343514919281,
+      "learning_rate": 4.518609569125034e-05,
+      "loss": 1.3655,
+      "step": 18221
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6030120849609375,
+      "learning_rate": 4.5185831350779805e-05,
+      "loss": 1.3217,
+      "step": 18222
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6623241901397705,
+      "learning_rate": 4.5185567010309285e-05,
+      "loss": 1.4328,
+      "step": 18223
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4885152578353882,
+      "learning_rate": 4.518530266983875e-05,
+      "loss": 1.2938,
+      "step": 18224
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8373069763183594,
+      "learning_rate": 4.518503832936823e-05,
+      "loss": 1.5134,
+      "step": 18225
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4064466953277588,
+      "learning_rate": 4.5184773988897706e-05,
+      "loss": 1.3719,
+      "step": 18226
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.063861846923828,
+      "learning_rate": 4.518450964842717e-05,
+      "loss": 1.3423,
+      "step": 18227
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6658576726913452,
+      "learning_rate": 4.518424530795665e-05,
+      "loss": 1.488,
+      "step": 18228
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4175976514816284,
+      "learning_rate": 4.518398096748612e-05,
+      "loss": 1.2249,
+      "step": 18229
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7399259805679321,
+      "learning_rate": 4.51837166270156e-05,
+      "loss": 1.41,
+      "step": 18230
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.928966999053955,
+      "learning_rate": 4.5183452286545074e-05,
+      "loss": 1.2276,
+      "step": 18231
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.734187364578247,
+      "learning_rate": 4.518318794607455e-05,
+      "loss": 1.3684,
+      "step": 18232
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.603754997253418,
+      "learning_rate": 4.518292360560402e-05,
+      "loss": 1.5132,
+      "step": 18233
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.497448444366455,
+      "learning_rate": 4.5182659265133494e-05,
+      "loss": 1.947,
+      "step": 18234
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7369550466537476,
+      "learning_rate": 4.518239492466297e-05,
+      "loss": 1.0983,
+      "step": 18235
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7385921478271484,
+      "learning_rate": 4.518213058419244e-05,
+      "loss": 1.3632,
+      "step": 18236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4595539569854736,
+      "learning_rate": 4.5181866243721915e-05,
+      "loss": 1.2029,
+      "step": 18237
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9528987407684326,
+      "learning_rate": 4.518160190325139e-05,
+      "loss": 1.2794,
+      "step": 18238
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.527294158935547,
+      "learning_rate": 4.518133756278086e-05,
+      "loss": 1.5704,
+      "step": 18239
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6185355186462402,
+      "learning_rate": 4.5181073222310336e-05,
+      "loss": 1.4195,
+      "step": 18240
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.043337345123291,
+      "learning_rate": 4.5180808881839816e-05,
+      "loss": 1.558,
+      "step": 18241
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5730208158493042,
+      "learning_rate": 4.518054454136928e-05,
+      "loss": 1.3694,
+      "step": 18242
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1037254333496094,
+      "learning_rate": 4.518028020089876e-05,
+      "loss": 1.5952,
+      "step": 18243
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.581632137298584,
+      "learning_rate": 4.518001586042823e-05,
+      "loss": 1.3279,
+      "step": 18244
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9137860536575317,
+      "learning_rate": 4.517975151995771e-05,
+      "loss": 1.4282,
+      "step": 18245
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5364679098129272,
+      "learning_rate": 4.5179487179487184e-05,
+      "loss": 1.3593,
+      "step": 18246
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.95315682888031,
+      "learning_rate": 4.517922283901666e-05,
+      "loss": 1.4774,
+      "step": 18247
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8262217044830322,
+      "learning_rate": 4.517895849854613e-05,
+      "loss": 1.3327,
+      "step": 18248
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6433840990066528,
+      "learning_rate": 4.5178694158075605e-05,
+      "loss": 1.4936,
+      "step": 18249
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1666910648345947,
+      "learning_rate": 4.517842981760508e-05,
+      "loss": 1.3941,
+      "step": 18250
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0524134635925293,
+      "learning_rate": 4.517816547713455e-05,
+      "loss": 1.5405,
+      "step": 18251
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6146844625473022,
+      "learning_rate": 4.5177901136664025e-05,
+      "loss": 1.2636,
+      "step": 18252
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9043500423431396,
+      "learning_rate": 4.51776367961935e-05,
+      "loss": 1.3458,
+      "step": 18253
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.040968179702759,
+      "learning_rate": 4.517737245572297e-05,
+      "loss": 1.5161,
+      "step": 18254
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.462801456451416,
+      "learning_rate": 4.5177108115252446e-05,
+      "loss": 1.202,
+      "step": 18255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8845876455307007,
+      "learning_rate": 4.5176843774781926e-05,
+      "loss": 1.3908,
+      "step": 18256
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8120431900024414,
+      "learning_rate": 4.517657943431139e-05,
+      "loss": 1.2981,
+      "step": 18257
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4132882356643677,
+      "learning_rate": 4.5176315093840874e-05,
+      "loss": 1.3591,
+      "step": 18258
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6518089771270752,
+      "learning_rate": 4.517605075337034e-05,
+      "loss": 1.2294,
+      "step": 18259
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9130905866622925,
+      "learning_rate": 4.517578641289982e-05,
+      "loss": 1.5639,
+      "step": 18260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9085912704467773,
+      "learning_rate": 4.5175522072429294e-05,
+      "loss": 1.5552,
+      "step": 18261
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8078536987304688,
+      "learning_rate": 4.517525773195877e-05,
+      "loss": 1.4921,
+      "step": 18262
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.774985432624817,
+      "learning_rate": 4.517499339148824e-05,
+      "loss": 1.3847,
+      "step": 18263
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7585357427597046,
+      "learning_rate": 4.517472905101771e-05,
+      "loss": 1.3174,
+      "step": 18264
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2271084785461426,
+      "learning_rate": 4.517446471054719e-05,
+      "loss": 1.4326,
+      "step": 18265
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.328184127807617,
+      "learning_rate": 4.517420037007666e-05,
+      "loss": 1.6577,
+      "step": 18266
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5021708011627197,
+      "learning_rate": 4.5173936029606136e-05,
+      "loss": 1.2161,
+      "step": 18267
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6018126010894775,
+      "learning_rate": 4.517367168913561e-05,
+      "loss": 1.3024,
+      "step": 18268
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9020054340362549,
+      "learning_rate": 4.517340734866508e-05,
+      "loss": 1.3619,
+      "step": 18269
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8174717426300049,
+      "learning_rate": 4.5173143008194556e-05,
+      "loss": 1.4044,
+      "step": 18270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5427707433700562,
+      "learning_rate": 4.517287866772404e-05,
+      "loss": 1.4133,
+      "step": 18271
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5894081592559814,
+      "learning_rate": 4.5172614327253504e-05,
+      "loss": 1.3374,
+      "step": 18272
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.747750997543335,
+      "learning_rate": 4.517234998678298e-05,
+      "loss": 1.3607,
+      "step": 18273
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5973784923553467,
+      "learning_rate": 4.517208564631245e-05,
+      "loss": 1.4202,
+      "step": 18274
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3779103755950928,
+      "learning_rate": 4.5171821305841924e-05,
+      "loss": 1.7577,
+      "step": 18275
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3190574645996094,
+      "learning_rate": 4.5171556965371405e-05,
+      "loss": 1.3239,
+      "step": 18276
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4614742994308472,
+      "learning_rate": 4.517129262490087e-05,
+      "loss": 1.1527,
+      "step": 18277
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.707140564918518,
+      "learning_rate": 4.517102828443035e-05,
+      "loss": 2.528,
+      "step": 18278
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2307379245758057,
+      "learning_rate": 4.517076394395982e-05,
+      "loss": 1.4566,
+      "step": 18279
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9153727293014526,
+      "learning_rate": 4.51704996034893e-05,
+      "loss": 1.3983,
+      "step": 18280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0511035919189453,
+      "learning_rate": 4.517023526301877e-05,
+      "loss": 1.4791,
+      "step": 18281
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.489140748977661,
+      "learning_rate": 4.5169970922548246e-05,
+      "loss": 1.5618,
+      "step": 18282
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.89822256565094,
+      "learning_rate": 4.516970658207772e-05,
+      "loss": 1.5106,
+      "step": 18283
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4111437797546387,
+      "learning_rate": 4.516944224160719e-05,
+      "loss": 1.5279,
+      "step": 18284
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6805088520050049,
+      "learning_rate": 4.516917790113667e-05,
+      "loss": 1.4106,
+      "step": 18285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.934371829032898,
+      "learning_rate": 4.516891356066614e-05,
+      "loss": 1.6279,
+      "step": 18286
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0055577754974365,
+      "learning_rate": 4.5168649220195614e-05,
+      "loss": 1.3968,
+      "step": 18287
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.94285249710083,
+      "learning_rate": 4.516838487972509e-05,
+      "loss": 1.3406,
+      "step": 18288
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.747812271118164,
+      "learning_rate": 4.516812053925456e-05,
+      "loss": 1.4333,
+      "step": 18289
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8642560243606567,
+      "learning_rate": 4.5167856198784035e-05,
+      "loss": 1.3999,
+      "step": 18290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7436778545379639,
+      "learning_rate": 4.5167591858313515e-05,
+      "loss": 1.4856,
+      "step": 18291
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8816630840301514,
+      "learning_rate": 4.516732751784298e-05,
+      "loss": 1.3664,
+      "step": 18292
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6159424781799316,
+      "learning_rate": 4.516706317737246e-05,
+      "loss": 1.2612,
+      "step": 18293
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7630711793899536,
+      "learning_rate": 4.516679883690193e-05,
+      "loss": 1.2113,
+      "step": 18294
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5482836961746216,
+      "learning_rate": 4.516653449643141e-05,
+      "loss": 1.3396,
+      "step": 18295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9076216220855713,
+      "learning_rate": 4.516627015596088e-05,
+      "loss": 1.3335,
+      "step": 18296
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7774920463562012,
+      "learning_rate": 4.5166005815490356e-05,
+      "loss": 1.3902,
+      "step": 18297
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8841793537139893,
+      "learning_rate": 4.516574147501983e-05,
+      "loss": 1.3388,
+      "step": 18298
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8608298301696777,
+      "learning_rate": 4.51654771345493e-05,
+      "loss": 1.3505,
+      "step": 18299
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2911972999572754,
+      "learning_rate": 4.516521279407878e-05,
+      "loss": 2.5198,
+      "step": 18300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.823427438735962,
+      "learning_rate": 4.516494845360825e-05,
+      "loss": 1.515,
+      "step": 18301
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.022991180419922,
+      "learning_rate": 4.5164684113137724e-05,
+      "loss": 1.2957,
+      "step": 18302
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.741703748703003,
+      "learning_rate": 4.51644197726672e-05,
+      "loss": 1.3918,
+      "step": 18303
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3110601902008057,
+      "learning_rate": 4.516415543219667e-05,
+      "loss": 1.6221,
+      "step": 18304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4649678468704224,
+      "learning_rate": 4.5163891091726145e-05,
+      "loss": 2.3527,
+      "step": 18305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.150902032852173,
+      "learning_rate": 4.5163626751255625e-05,
+      "loss": 1.5813,
+      "step": 18306
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.436087965965271,
+      "learning_rate": 4.516336241078509e-05,
+      "loss": 1.3507,
+      "step": 18307
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.74179744720459,
+      "learning_rate": 4.5163098070314566e-05,
+      "loss": 1.4435,
+      "step": 18308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5823758840560913,
+      "learning_rate": 4.516283372984404e-05,
+      "loss": 1.3491,
+      "step": 18309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.654765248298645,
+      "learning_rate": 4.516256938937351e-05,
+      "loss": 1.4571,
+      "step": 18310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.016963481903076,
+      "learning_rate": 4.516230504890299e-05,
+      "loss": 1.3145,
+      "step": 18311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6341270208358765,
+      "learning_rate": 4.516204070843246e-05,
+      "loss": 1.3135,
+      "step": 18312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6901549100875854,
+      "learning_rate": 4.516177636796194e-05,
+      "loss": 1.3052,
+      "step": 18313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4746919870376587,
+      "learning_rate": 4.516151202749141e-05,
+      "loss": 2.3791,
+      "step": 18314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.460600733757019,
+      "learning_rate": 4.516124768702089e-05,
+      "loss": 1.2654,
+      "step": 18315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7895179986953735,
+      "learning_rate": 4.516098334655036e-05,
+      "loss": 1.307,
+      "step": 18316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.744261622428894,
+      "learning_rate": 4.5160719006079835e-05,
+      "loss": 2.1049,
+      "step": 18317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6437110900878906,
+      "learning_rate": 4.516045466560931e-05,
+      "loss": 1.4031,
+      "step": 18318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5846621990203857,
+      "learning_rate": 4.516019032513878e-05,
+      "loss": 1.3249,
+      "step": 18319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5261497497558594,
+      "learning_rate": 4.5159925984668255e-05,
+      "loss": 1.4664,
+      "step": 18320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.866227149963379,
+      "learning_rate": 4.515966164419773e-05,
+      "loss": 1.4292,
+      "step": 18321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4252182245254517,
+      "learning_rate": 4.51593973037272e-05,
+      "loss": 1.2421,
+      "step": 18322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6248571872711182,
+      "learning_rate": 4.5159132963256676e-05,
+      "loss": 1.3633,
+      "step": 18323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6685787439346313,
+      "learning_rate": 4.515886862278615e-05,
+      "loss": 1.4584,
+      "step": 18324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0652148723602295,
+      "learning_rate": 4.515860428231562e-05,
+      "loss": 1.2837,
+      "step": 18325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4097787141799927,
+      "learning_rate": 4.5158339941845103e-05,
+      "loss": 1.3054,
+      "step": 18326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0414986610412598,
+      "learning_rate": 4.515807560137457e-05,
+      "loss": 1.4025,
+      "step": 18327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8215858936309814,
+      "learning_rate": 4.515781126090405e-05,
+      "loss": 1.4656,
+      "step": 18328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5245355367660522,
+      "learning_rate": 4.515754692043352e-05,
+      "loss": 2.0174,
+      "step": 18329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4386019706726074,
+      "learning_rate": 4.5157282579963e-05,
+      "loss": 1.2739,
+      "step": 18330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7652021646499634,
+      "learning_rate": 4.515701823949247e-05,
+      "loss": 1.6195,
+      "step": 18331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.722421407699585,
+      "learning_rate": 4.5156753899021945e-05,
+      "loss": 1.4053,
+      "step": 18332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.136887311935425,
+      "learning_rate": 4.515648955855142e-05,
+      "loss": 1.4217,
+      "step": 18333
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6263055801391602,
+      "learning_rate": 4.515622521808089e-05,
+      "loss": 1.4195,
+      "step": 18334
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.341496229171753,
+      "learning_rate": 4.5155960877610366e-05,
+      "loss": 1.4567,
+      "step": 18335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.130993604660034,
+      "learning_rate": 4.515569653713983e-05,
+      "loss": 1.379,
+      "step": 18336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.830114722251892,
+      "learning_rate": 4.515543219666931e-05,
+      "loss": 1.379,
+      "step": 18337
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.595117449760437,
+      "learning_rate": 4.5155167856198786e-05,
+      "loss": 1.2029,
+      "step": 18338
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.699713945388794,
+      "learning_rate": 4.515490351572826e-05,
+      "loss": 1.4858,
+      "step": 18339
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7991043329238892,
+      "learning_rate": 4.5154639175257733e-05,
+      "loss": 1.5494,
+      "step": 18340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9847999811172485,
+      "learning_rate": 4.515437483478721e-05,
+      "loss": 1.3407,
+      "step": 18341
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3689801692962646,
+      "learning_rate": 4.515411049431668e-05,
+      "loss": 1.488,
+      "step": 18342
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4213076829910278,
+      "learning_rate": 4.515384615384616e-05,
+      "loss": 1.2919,
+      "step": 18343
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8943004608154297,
+      "learning_rate": 4.515358181337563e-05,
+      "loss": 1.2703,
+      "step": 18344
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0378875732421875,
+      "learning_rate": 4.51533174729051e-05,
+      "loss": 1.7775,
+      "step": 18345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7656632661819458,
+      "learning_rate": 4.5153053132434575e-05,
+      "loss": 1.2432,
+      "step": 18346
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0891339778900146,
+      "learning_rate": 4.515278879196405e-05,
+      "loss": 1.4565,
+      "step": 18347
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9537485837936401,
+      "learning_rate": 4.515252445149353e-05,
+      "loss": 1.5355,
+      "step": 18348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9670805931091309,
+      "learning_rate": 4.5152260111022996e-05,
+      "loss": 1.3193,
+      "step": 18349
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9925764799118042,
+      "learning_rate": 4.5151995770552476e-05,
+      "loss": 1.4527,
+      "step": 18350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.337052583694458,
+      "learning_rate": 4.515173143008194e-05,
+      "loss": 1.298,
+      "step": 18351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5436474084854126,
+      "learning_rate": 4.515146708961142e-05,
+      "loss": 1.2492,
+      "step": 18352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9071025848388672,
+      "learning_rate": 4.51512027491409e-05,
+      "loss": 1.3945,
+      "step": 18353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.076059103012085,
+      "learning_rate": 4.515093840867037e-05,
+      "loss": 1.4677,
+      "step": 18354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.752598762512207,
+      "learning_rate": 4.5150674068199844e-05,
+      "loss": 1.4184,
+      "step": 18355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7089602947235107,
+      "learning_rate": 4.515040972772932e-05,
+      "loss": 1.4837,
+      "step": 18356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7120805978775024,
+      "learning_rate": 4.515014538725879e-05,
+      "loss": 1.3048,
+      "step": 18357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.176778793334961,
+      "learning_rate": 4.5149881046788265e-05,
+      "loss": 1.2575,
+      "step": 18358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.769714593887329,
+      "learning_rate": 4.514961670631774e-05,
+      "loss": 1.3258,
+      "step": 18359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.96962571144104,
+      "learning_rate": 4.514935236584721e-05,
+      "loss": 1.4267,
+      "step": 18360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6399922370910645,
+      "learning_rate": 4.5149088025376685e-05,
+      "loss": 1.2662,
+      "step": 18361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0429608821868896,
+      "learning_rate": 4.514882368490616e-05,
+      "loss": 1.2246,
+      "step": 18362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8020175695419312,
+      "learning_rate": 4.514855934443564e-05,
+      "loss": 1.3203,
+      "step": 18363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.109511613845825,
+      "learning_rate": 4.5148295003965106e-05,
+      "loss": 1.4753,
+      "step": 18364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9502003192901611,
+      "learning_rate": 4.5148030663494586e-05,
+      "loss": 1.4791,
+      "step": 18365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6589301824569702,
+      "learning_rate": 4.514776632302405e-05,
+      "loss": 1.3884,
+      "step": 18366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9153285026550293,
+      "learning_rate": 4.5147501982553533e-05,
+      "loss": 1.4384,
+      "step": 18367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9475529193878174,
+      "learning_rate": 4.514723764208301e-05,
+      "loss": 1.3743,
+      "step": 18368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0589287281036377,
+      "learning_rate": 4.514697330161248e-05,
+      "loss": 1.6056,
+      "step": 18369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7151998281478882,
+      "learning_rate": 4.5146708961141954e-05,
+      "loss": 1.3019,
+      "step": 18370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.243455410003662,
+      "learning_rate": 4.514644462067142e-05,
+      "loss": 1.4924,
+      "step": 18371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.624691128730774,
+      "learning_rate": 4.51461802802009e-05,
+      "loss": 1.3913,
+      "step": 18372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.823298692703247,
+      "learning_rate": 4.5145915939730375e-05,
+      "loss": 1.3164,
+      "step": 18373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4117143154144287,
+      "learning_rate": 4.514565159925985e-05,
+      "loss": 1.3012,
+      "step": 18374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.70903742313385,
+      "learning_rate": 4.514538725878932e-05,
+      "loss": 1.2905,
+      "step": 18375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7302590608596802,
+      "learning_rate": 4.5145122918318796e-05,
+      "loss": 1.2667,
+      "step": 18376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8157517910003662,
+      "learning_rate": 4.514485857784827e-05,
+      "loss": 1.2933,
+      "step": 18377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4557636976242065,
+      "learning_rate": 4.514459423737775e-05,
+      "loss": 1.3161,
+      "step": 18378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3830742835998535,
+      "learning_rate": 4.5144329896907216e-05,
+      "loss": 1.2688,
+      "step": 18379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6061193943023682,
+      "learning_rate": 4.514406555643669e-05,
+      "loss": 1.1133,
+      "step": 18380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4843189716339111,
+      "learning_rate": 4.5143801215966163e-05,
+      "loss": 1.584,
+      "step": 18381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9051610231399536,
+      "learning_rate": 4.514353687549564e-05,
+      "loss": 2.076,
+      "step": 18382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9208884239196777,
+      "learning_rate": 4.514327253502512e-05,
+      "loss": 1.2167,
+      "step": 18383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.588395357131958,
+      "learning_rate": 4.5143008194554584e-05,
+      "loss": 1.285,
+      "step": 18384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8524290323257446,
+      "learning_rate": 4.5142743854084064e-05,
+      "loss": 1.3533,
+      "step": 18385
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7102775573730469,
+      "learning_rate": 4.514247951361353e-05,
+      "loss": 1.3716,
+      "step": 18386
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6567436456680298,
+      "learning_rate": 4.514221517314301e-05,
+      "loss": 1.3764,
+      "step": 18387
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7679901123046875,
+      "learning_rate": 4.5141950832672485e-05,
+      "loss": 1.4049,
+      "step": 18388
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9221251010894775,
+      "learning_rate": 4.514168649220196e-05,
+      "loss": 1.3733,
+      "step": 18389
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6150360107421875,
+      "learning_rate": 4.514142215173143e-05,
+      "loss": 1.324,
+      "step": 18390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5253671407699585,
+      "learning_rate": 4.5141157811260906e-05,
+      "loss": 1.2492,
+      "step": 18391
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9838072061538696,
+      "learning_rate": 4.514089347079038e-05,
+      "loss": 1.4349,
+      "step": 18392
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7185064554214478,
+      "learning_rate": 4.514062913031985e-05,
+      "loss": 1.261,
+      "step": 18393
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6665148735046387,
+      "learning_rate": 4.5140364789849327e-05,
+      "loss": 1.3186,
+      "step": 18394
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6715610027313232,
+      "learning_rate": 4.51401004493788e-05,
+      "loss": 1.2111,
+      "step": 18395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8268800973892212,
+      "learning_rate": 4.5139836108908274e-05,
+      "loss": 1.4699,
+      "step": 18396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4206626415252686,
+      "learning_rate": 4.513957176843775e-05,
+      "loss": 2.1581,
+      "step": 18397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0212461948394775,
+      "learning_rate": 4.513930742796723e-05,
+      "loss": 1.3988,
+      "step": 18398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.33299720287323,
+      "learning_rate": 4.5139043087496694e-05,
+      "loss": 1.3359,
+      "step": 18399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8550286293029785,
+      "learning_rate": 4.5138778747026175e-05,
+      "loss": 1.3811,
+      "step": 18400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6411516666412354,
+      "learning_rate": 4.513851440655564e-05,
+      "loss": 1.5322,
+      "step": 18401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6108596324920654,
+      "learning_rate": 4.513825006608512e-05,
+      "loss": 1.4334,
+      "step": 18402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7511476278305054,
+      "learning_rate": 4.5137985725614596e-05,
+      "loss": 1.3646,
+      "step": 18403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.274582624435425,
+      "learning_rate": 4.513772138514407e-05,
+      "loss": 1.4273,
+      "step": 18404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8106969594955444,
+      "learning_rate": 4.513745704467354e-05,
+      "loss": 1.4105,
+      "step": 18405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4257724285125732,
+      "learning_rate": 4.5137192704203016e-05,
+      "loss": 1.4572,
+      "step": 18406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0151751041412354,
+      "learning_rate": 4.513692836373249e-05,
+      "loss": 1.3298,
+      "step": 18407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8583356142044067,
+      "learning_rate": 4.513666402326196e-05,
+      "loss": 1.4884,
+      "step": 18408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4724937677383423,
+      "learning_rate": 4.513639968279144e-05,
+      "loss": 1.9641,
+      "step": 18409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4935123920440674,
+      "learning_rate": 4.513613534232091e-05,
+      "loss": 1.4711,
+      "step": 18410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.519585371017456,
+      "learning_rate": 4.5135871001850384e-05,
+      "loss": 1.226,
+      "step": 18411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.425428628921509,
+      "learning_rate": 4.513560666137986e-05,
+      "loss": 1.3862,
+      "step": 18412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5069198608398438,
+      "learning_rate": 4.513534232090934e-05,
+      "loss": 1.5796,
+      "step": 18413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6834665536880493,
+      "learning_rate": 4.5135077980438805e-05,
+      "loss": 1.1802,
+      "step": 18414
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3492645025253296,
+      "learning_rate": 4.5134813639968285e-05,
+      "loss": 1.2467,
+      "step": 18415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2521395683288574,
+      "learning_rate": 4.513454929949775e-05,
+      "loss": 1.4406,
+      "step": 18416
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6716578006744385,
+      "learning_rate": 4.5134284959027226e-05,
+      "loss": 1.5288,
+      "step": 18417
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6429270505905151,
+      "learning_rate": 4.5134020618556706e-05,
+      "loss": 1.2893,
+      "step": 18418
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3897582292556763,
+      "learning_rate": 4.513375627808617e-05,
+      "loss": 1.3508,
+      "step": 18419
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2894275188446045,
+      "learning_rate": 4.513349193761565e-05,
+      "loss": 1.6365,
+      "step": 18420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.398270606994629,
+      "learning_rate": 4.513322759714512e-05,
+      "loss": 1.6786,
+      "step": 18421
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0693888664245605,
+      "learning_rate": 4.51329632566746e-05,
+      "loss": 1.3753,
+      "step": 18422
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.603356122970581,
+      "learning_rate": 4.5132698916204074e-05,
+      "loss": 1.5228,
+      "step": 18423
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6321172714233398,
+      "learning_rate": 4.513243457573355e-05,
+      "loss": 1.4154,
+      "step": 18424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4970471858978271,
+      "learning_rate": 4.513217023526302e-05,
+      "loss": 1.277,
+      "step": 18425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.244422674179077,
+      "learning_rate": 4.5131905894792494e-05,
+      "loss": 1.5942,
+      "step": 18426
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7297605276107788,
+      "learning_rate": 4.513164155432197e-05,
+      "loss": 1.4584,
+      "step": 18427
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6955173015594482,
+      "learning_rate": 4.513137721385144e-05,
+      "loss": 1.2606,
+      "step": 18428
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7768601179122925,
+      "learning_rate": 4.5131112873380915e-05,
+      "loss": 1.2929,
+      "step": 18429
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9747447967529297,
+      "learning_rate": 4.513084853291039e-05,
+      "loss": 1.4142,
+      "step": 18430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.659690260887146,
+      "learning_rate": 4.513058419243986e-05,
+      "loss": 1.4375,
+      "step": 18431
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5584872961044312,
+      "learning_rate": 4.5130319851969336e-05,
+      "loss": 1.3711,
+      "step": 18432
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5205157995224,
+      "learning_rate": 4.5130055511498816e-05,
+      "loss": 1.4108,
+      "step": 18433
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5583935976028442,
+      "learning_rate": 4.512979117102828e-05,
+      "loss": 1.284,
+      "step": 18434
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0227134227752686,
+      "learning_rate": 4.512952683055776e-05,
+      "loss": 1.6255,
+      "step": 18435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8300246000289917,
+      "learning_rate": 4.512926249008723e-05,
+      "loss": 1.389,
+      "step": 18436
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1805474758148193,
+      "learning_rate": 4.512899814961671e-05,
+      "loss": 1.3222,
+      "step": 18437
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8892110586166382,
+      "learning_rate": 4.5128733809146184e-05,
+      "loss": 1.4262,
+      "step": 18438
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7227836847305298,
+      "learning_rate": 4.512846946867566e-05,
+      "loss": 1.3646,
+      "step": 18439
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5816638469696045,
+      "learning_rate": 4.512820512820513e-05,
+      "loss": 1.3476,
+      "step": 18440
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8778491020202637,
+      "learning_rate": 4.5127940787734605e-05,
+      "loss": 1.2675,
+      "step": 18441
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2422869205474854,
+      "learning_rate": 4.512767644726408e-05,
+      "loss": 1.4556,
+      "step": 18442
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7327916622161865,
+      "learning_rate": 4.512741210679355e-05,
+      "loss": 1.3576,
+      "step": 18443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9230949878692627,
+      "learning_rate": 4.5127147766323025e-05,
+      "loss": 1.5492,
+      "step": 18444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.609743595123291,
+      "learning_rate": 4.51268834258525e-05,
+      "loss": 1.3787,
+      "step": 18445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.937086820602417,
+      "learning_rate": 4.512661908538197e-05,
+      "loss": 1.4975,
+      "step": 18446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.314990282058716,
+      "learning_rate": 4.5126354744911446e-05,
+      "loss": 1.4765,
+      "step": 18447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2280831336975098,
+      "learning_rate": 4.5126090404440927e-05,
+      "loss": 1.3911,
+      "step": 18448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2778093814849854,
+      "learning_rate": 4.512582606397039e-05,
+      "loss": 1.4151,
+      "step": 18449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2372567653656006,
+      "learning_rate": 4.5125561723499874e-05,
+      "loss": 1.508,
+      "step": 18450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6477723121643066,
+      "learning_rate": 4.512529738302934e-05,
+      "loss": 1.3009,
+      "step": 18451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.123216152191162,
+      "learning_rate": 4.512503304255882e-05,
+      "loss": 1.4065,
+      "step": 18452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5745946168899536,
+      "learning_rate": 4.5124768702088294e-05,
+      "loss": 1.9817,
+      "step": 18453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5785925388336182,
+      "learning_rate": 4.512450436161776e-05,
+      "loss": 1.2286,
+      "step": 18454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9122675657272339,
+      "learning_rate": 4.512424002114724e-05,
+      "loss": 1.4473,
+      "step": 18455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3160338401794434,
+      "learning_rate": 4.512397568067671e-05,
+      "loss": 1.4576,
+      "step": 18456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1713707447052,
+      "learning_rate": 4.512371134020619e-05,
+      "loss": 1.6238,
+      "step": 18457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.659468173980713,
+      "learning_rate": 4.512344699973566e-05,
+      "loss": 1.4093,
+      "step": 18458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.855085015296936,
+      "learning_rate": 4.5123182659265136e-05,
+      "loss": 2.3165,
+      "step": 18459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5199310779571533,
+      "learning_rate": 4.512291831879461e-05,
+      "loss": 2.0925,
+      "step": 18460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4703220129013062,
+      "learning_rate": 4.512265397832408e-05,
+      "loss": 1.0918,
+      "step": 18461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7885451316833496,
+      "learning_rate": 4.5122389637853557e-05,
+      "loss": 1.5009,
+      "step": 18462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9350851774215698,
+      "learning_rate": 4.512212529738303e-05,
+      "loss": 1.4184,
+      "step": 18463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5574116706848145,
+      "learning_rate": 4.5121860956912504e-05,
+      "loss": 1.8379,
+      "step": 18464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1224446296691895,
+      "learning_rate": 4.512159661644198e-05,
+      "loss": 2.2138,
+      "step": 18465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4134857654571533,
+      "learning_rate": 4.512133227597145e-05,
+      "loss": 1.5982,
+      "step": 18466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7529007196426392,
+      "learning_rate": 4.5121067935500924e-05,
+      "loss": 1.4596,
+      "step": 18467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4807666540145874,
+      "learning_rate": 4.5120803595030405e-05,
+      "loss": 1.4013,
+      "step": 18468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.795635461807251,
+      "learning_rate": 4.512053925455987e-05,
+      "loss": 1.4144,
+      "step": 18469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8523143529891968,
+      "learning_rate": 4.512027491408935e-05,
+      "loss": 1.3603,
+      "step": 18470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6544991731643677,
+      "learning_rate": 4.512001057361882e-05,
+      "loss": 1.3273,
+      "step": 18471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6027876138687134,
+      "learning_rate": 4.51197462331483e-05,
+      "loss": 1.3761,
+      "step": 18472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8433654308319092,
+      "learning_rate": 4.511948189267777e-05,
+      "loss": 1.3038,
+      "step": 18473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.940353512763977,
+      "learning_rate": 4.5119217552207246e-05,
+      "loss": 1.3811,
+      "step": 18474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5983628034591675,
+      "learning_rate": 4.511895321173672e-05,
+      "loss": 1.3306,
+      "step": 18475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8004989624023438,
+      "learning_rate": 4.511868887126619e-05,
+      "loss": 1.4152,
+      "step": 18476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5754969120025635,
+      "learning_rate": 4.511842453079567e-05,
+      "loss": 1.4039,
+      "step": 18477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.228856325149536,
+      "learning_rate": 4.511816019032514e-05,
+      "loss": 1.4431,
+      "step": 18478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.003941059112549,
+      "learning_rate": 4.5117895849854614e-05,
+      "loss": 1.4969,
+      "step": 18479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.764015793800354,
+      "learning_rate": 4.511763150938409e-05,
+      "loss": 1.2112,
+      "step": 18480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7657285928726196,
+      "learning_rate": 4.511736716891356e-05,
+      "loss": 1.4536,
+      "step": 18481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.123579502105713,
+      "learning_rate": 4.5117102828443035e-05,
+      "loss": 1.4162,
+      "step": 18482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7074097394943237,
+      "learning_rate": 4.5116838487972515e-05,
+      "loss": 1.3744,
+      "step": 18483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9654204845428467,
+      "learning_rate": 4.511657414750198e-05,
+      "loss": 1.5252,
+      "step": 18484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4762548208236694,
+      "learning_rate": 4.511630980703146e-05,
+      "loss": 2.2092,
+      "step": 18485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4041881561279297,
+      "learning_rate": 4.511604546656093e-05,
+      "loss": 1.2906,
+      "step": 18486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7642980813980103,
+      "learning_rate": 4.511578112609041e-05,
+      "loss": 1.3593,
+      "step": 18487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2402291297912598,
+      "learning_rate": 4.511551678561988e-05,
+      "loss": 1.4598,
+      "step": 18488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.517953634262085,
+      "learning_rate": 4.511525244514935e-05,
+      "loss": 1.521,
+      "step": 18489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7003769874572754,
+      "learning_rate": 4.511498810467883e-05,
+      "loss": 1.1938,
+      "step": 18490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6509454250335693,
+      "learning_rate": 4.51147237642083e-05,
+      "loss": 1.4488,
+      "step": 18491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.839773178100586,
+      "learning_rate": 4.511445942373778e-05,
+      "loss": 1.2143,
+      "step": 18492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.809415340423584,
+      "learning_rate": 4.511419508326725e-05,
+      "loss": 1.3408,
+      "step": 18493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8150584697723389,
+      "learning_rate": 4.5113930742796724e-05,
+      "loss": 1.3802,
+      "step": 18494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.369972586631775,
+      "learning_rate": 4.51136664023262e-05,
+      "loss": 1.2084,
+      "step": 18495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8279402256011963,
+      "learning_rate": 4.511340206185567e-05,
+      "loss": 1.3708,
+      "step": 18496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5655603408813477,
+      "learning_rate": 4.5113137721385145e-05,
+      "loss": 2.0309,
+      "step": 18497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5624510049819946,
+      "learning_rate": 4.511287338091462e-05,
+      "loss": 1.1824,
+      "step": 18498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.776265263557434,
+      "learning_rate": 4.511260904044409e-05,
+      "loss": 2.541,
+      "step": 18499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1437184810638428,
+      "learning_rate": 4.5112344699973566e-05,
+      "loss": 1.5262,
+      "step": 18500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.003270387649536,
+      "learning_rate": 4.511208035950304e-05,
+      "loss": 1.4702,
+      "step": 18501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5660277605056763,
+      "learning_rate": 4.511181601903251e-05,
+      "loss": 1.346,
+      "step": 18502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.013031005859375,
+      "learning_rate": 4.511155167856199e-05,
+      "loss": 1.4352,
+      "step": 18503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6714304685592651,
+      "learning_rate": 4.511128733809146e-05,
+      "loss": 1.3309,
+      "step": 18504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8293040990829468,
+      "learning_rate": 4.511102299762094e-05,
+      "loss": 1.4128,
+      "step": 18505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.93440842628479,
+      "learning_rate": 4.511075865715041e-05,
+      "loss": 1.6072,
+      "step": 18506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.823793649673462,
+      "learning_rate": 4.511049431667989e-05,
+      "loss": 1.4128,
+      "step": 18507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4731758832931519,
+      "learning_rate": 4.511022997620936e-05,
+      "loss": 1.3009,
+      "step": 18508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6737818717956543,
+      "learning_rate": 4.5109965635738835e-05,
+      "loss": 1.4401,
+      "step": 18509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.511646032333374,
+      "learning_rate": 4.510970129526831e-05,
+      "loss": 1.381,
+      "step": 18510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.132143020629883,
+      "learning_rate": 4.510943695479778e-05,
+      "loss": 1.3829,
+      "step": 18511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5409313440322876,
+      "learning_rate": 4.5109172614327255e-05,
+      "loss": 1.3535,
+      "step": 18512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.616391181945801,
+      "learning_rate": 4.510890827385673e-05,
+      "loss": 1.4898,
+      "step": 18513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.717569351196289,
+      "learning_rate": 4.51086439333862e-05,
+      "loss": 1.4634,
+      "step": 18514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5958548784255981,
+      "learning_rate": 4.5108379592915676e-05,
+      "loss": 1.3103,
+      "step": 18515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7372572422027588,
+      "learning_rate": 4.510811525244515e-05,
+      "loss": 1.2553,
+      "step": 18516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.014436960220337,
+      "learning_rate": 4.510785091197462e-05,
+      "loss": 1.4609,
+      "step": 18517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.497347354888916,
+      "learning_rate": 4.5107586571504104e-05,
+      "loss": 1.3404,
+      "step": 18518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6640114784240723,
+      "learning_rate": 4.510732223103357e-05,
+      "loss": 1.3984,
+      "step": 18519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9485149383544922,
+      "learning_rate": 4.510705789056305e-05,
+      "loss": 1.439,
+      "step": 18520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.88258695602417,
+      "learning_rate": 4.510679355009252e-05,
+      "loss": 1.4186,
+      "step": 18521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.606539487838745,
+      "learning_rate": 4.5106529209622e-05,
+      "loss": 1.4527,
+      "step": 18522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.620497226715088,
+      "learning_rate": 4.510626486915147e-05,
+      "loss": 2.2836,
+      "step": 18523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8075062036514282,
+      "learning_rate": 4.5106000528680945e-05,
+      "loss": 1.3046,
+      "step": 18524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1530091762542725,
+      "learning_rate": 4.510573618821042e-05,
+      "loss": 1.4172,
+      "step": 18525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5437891483306885,
+      "learning_rate": 4.5105471847739885e-05,
+      "loss": 1.4919,
+      "step": 18526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6178209781646729,
+      "learning_rate": 4.5105207507269366e-05,
+      "loss": 1.4235,
+      "step": 18527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7120370864868164,
+      "learning_rate": 4.510494316679884e-05,
+      "loss": 1.5903,
+      "step": 18528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6217180490493774,
+      "learning_rate": 4.510467882632831e-05,
+      "loss": 1.2417,
+      "step": 18529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7632862329483032,
+      "learning_rate": 4.5104414485857786e-05,
+      "loss": 1.4661,
+      "step": 18530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9584881067276,
+      "learning_rate": 4.510415014538726e-05,
+      "loss": 1.388,
+      "step": 18531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9560190439224243,
+      "learning_rate": 4.5103885804916734e-05,
+      "loss": 1.4057,
+      "step": 18532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8914644718170166,
+      "learning_rate": 4.5103621464446214e-05,
+      "loss": 1.4344,
+      "step": 18533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6474268436431885,
+      "learning_rate": 4.510335712397568e-05,
+      "loss": 1.3993,
+      "step": 18534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5851123332977295,
+      "learning_rate": 4.5103092783505154e-05,
+      "loss": 1.3427,
+      "step": 18535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0713164806365967,
+      "learning_rate": 4.510282844303463e-05,
+      "loss": 1.5123,
+      "step": 18536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2406203746795654,
+      "learning_rate": 4.51025641025641e-05,
+      "loss": 1.5012,
+      "step": 18537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.864632487297058,
+      "learning_rate": 4.510229976209358e-05,
+      "loss": 1.4738,
+      "step": 18538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2936463356018066,
+      "learning_rate": 4.510203542162305e-05,
+      "loss": 1.3408,
+      "step": 18539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.777944803237915,
+      "learning_rate": 4.510177108115253e-05,
+      "loss": 1.4414,
+      "step": 18540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.226203203201294,
+      "learning_rate": 4.5101506740681996e-05,
+      "loss": 1.5433,
+      "step": 18541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4524753093719482,
+      "learning_rate": 4.5101242400211476e-05,
+      "loss": 1.6304,
+      "step": 18542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7916009426116943,
+      "learning_rate": 4.510097805974095e-05,
+      "loss": 1.4927,
+      "step": 18543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6760011911392212,
+      "learning_rate": 4.510071371927042e-05,
+      "loss": 1.4329,
+      "step": 18544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8417954444885254,
+      "learning_rate": 4.51004493787999e-05,
+      "loss": 1.4589,
+      "step": 18545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5742558240890503,
+      "learning_rate": 4.510018503832937e-05,
+      "loss": 1.1568,
+      "step": 18546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.552603840827942,
+      "learning_rate": 4.5099920697858844e-05,
+      "loss": 1.1568,
+      "step": 18547
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.813782811164856,
+      "learning_rate": 4.509965635738832e-05,
+      "loss": 1.5041,
+      "step": 18548
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9379032850265503,
+      "learning_rate": 4.509939201691779e-05,
+      "loss": 1.5233,
+      "step": 18549
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4028241634368896,
+      "learning_rate": 4.5099127676447265e-05,
+      "loss": 1.1188,
+      "step": 18550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5943562984466553,
+      "learning_rate": 4.509886333597674e-05,
+      "loss": 1.3646,
+      "step": 18551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6723734140396118,
+      "learning_rate": 4.509859899550621e-05,
+      "loss": 1.3637,
+      "step": 18552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.775312900543213,
+      "learning_rate": 4.509833465503569e-05,
+      "loss": 1.3423,
+      "step": 18553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.619952917098999,
+      "learning_rate": 4.509807031456516e-05,
+      "loss": 1.3188,
+      "step": 18554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8434869050979614,
+      "learning_rate": 4.509780597409464e-05,
+      "loss": 1.4726,
+      "step": 18555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6156762838363647,
+      "learning_rate": 4.5097541633624106e-05,
+      "loss": 1.2511,
+      "step": 18556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8526866436004639,
+      "learning_rate": 4.5097277293153586e-05,
+      "loss": 1.3436,
+      "step": 18557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6834750175476074,
+      "learning_rate": 4.509701295268306e-05,
+      "loss": 1.3264,
+      "step": 18558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8584264516830444,
+      "learning_rate": 4.5096748612212534e-05,
+      "loss": 1.5259,
+      "step": 18559
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.551589012145996,
+      "learning_rate": 4.509648427174201e-05,
+      "loss": 2.2806,
+      "step": 18560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.096714496612549,
+      "learning_rate": 4.5096219931271474e-05,
+      "loss": 1.4771,
+      "step": 18561
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5934340953826904,
+      "learning_rate": 4.5095955590800954e-05,
+      "loss": 1.3024,
+      "step": 18562
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7843103408813477,
+      "learning_rate": 4.509569125033043e-05,
+      "loss": 1.6436,
+      "step": 18563
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8102102279663086,
+      "learning_rate": 4.50954269098599e-05,
+      "loss": 1.3136,
+      "step": 18564
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5372560024261475,
+      "learning_rate": 4.5095162569389375e-05,
+      "loss": 1.2252,
+      "step": 18565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.774648666381836,
+      "learning_rate": 4.509489822891885e-05,
+      "loss": 1.3783,
+      "step": 18566
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2378531694412231,
+      "learning_rate": 4.509463388844832e-05,
+      "loss": 1.124,
+      "step": 18567
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5592215061187744,
+      "learning_rate": 4.50943695479778e-05,
+      "loss": 1.3463,
+      "step": 18568
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9656157493591309,
+      "learning_rate": 4.509410520750727e-05,
+      "loss": 1.3666,
+      "step": 18569
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0865275859832764,
+      "learning_rate": 4.509384086703675e-05,
+      "loss": 1.5321,
+      "step": 18570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2731883525848389,
+      "learning_rate": 4.5093576526566216e-05,
+      "loss": 2.3206,
+      "step": 18571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5670890808105469,
+      "learning_rate": 4.509331218609569e-05,
+      "loss": 1.3703,
+      "step": 18572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7768828868865967,
+      "learning_rate": 4.509304784562517e-05,
+      "loss": 1.3325,
+      "step": 18573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.188929319381714,
+      "learning_rate": 4.509278350515464e-05,
+      "loss": 1.326,
+      "step": 18574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7010225057601929,
+      "learning_rate": 4.509251916468412e-05,
+      "loss": 1.3893,
+      "step": 18575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4733189344406128,
+      "learning_rate": 4.5092254824213584e-05,
+      "loss": 1.3095,
+      "step": 18576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.591464638710022,
+      "learning_rate": 4.5091990483743065e-05,
+      "loss": 1.2393,
+      "step": 18577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6543153524398804,
+      "learning_rate": 4.509172614327254e-05,
+      "loss": 1.4239,
+      "step": 18578
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.955636739730835,
+      "learning_rate": 4.509146180280201e-05,
+      "loss": 2.0788,
+      "step": 18579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1088786125183105,
+      "learning_rate": 4.5091197462331485e-05,
+      "loss": 1.5885,
+      "step": 18580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.780792236328125,
+      "learning_rate": 4.509093312186096e-05,
+      "loss": 1.235,
+      "step": 18581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7837992906570435,
+      "learning_rate": 4.509066878139043e-05,
+      "loss": 1.475,
+      "step": 18582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.529025077819824,
+      "learning_rate": 4.5090404440919906e-05,
+      "loss": 1.4802,
+      "step": 18583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2580432891845703,
+      "learning_rate": 4.509014010044938e-05,
+      "loss": 1.4214,
+      "step": 18584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0280981063842773,
+      "learning_rate": 4.508987575997885e-05,
+      "loss": 1.4108,
+      "step": 18585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.553856611251831,
+      "learning_rate": 4.508961141950833e-05,
+      "loss": 1.2638,
+      "step": 18586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9925506114959717,
+      "learning_rate": 4.50893470790378e-05,
+      "loss": 1.565,
+      "step": 18587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.654484748840332,
+      "learning_rate": 4.508908273856728e-05,
+      "loss": 1.1477,
+      "step": 18588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4463675022125244,
+      "learning_rate": 4.508881839809675e-05,
+      "loss": 1.4965,
+      "step": 18589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.268073081970215,
+      "learning_rate": 4.508855405762623e-05,
+      "loss": 1.6032,
+      "step": 18590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6189491748809814,
+      "learning_rate": 4.5088289717155695e-05,
+      "loss": 1.5314,
+      "step": 18591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.679176688194275,
+      "learning_rate": 4.5088025376685175e-05,
+      "loss": 1.292,
+      "step": 18592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.027301549911499,
+      "learning_rate": 4.508776103621465e-05,
+      "loss": 1.5311,
+      "step": 18593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.938088059425354,
+      "learning_rate": 4.508749669574412e-05,
+      "loss": 1.3524,
+      "step": 18594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5749760866165161,
+      "learning_rate": 4.5087232355273596e-05,
+      "loss": 1.1904,
+      "step": 18595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4044597148895264,
+      "learning_rate": 4.508696801480307e-05,
+      "loss": 2.1937,
+      "step": 18596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6500537395477295,
+      "learning_rate": 4.508670367433254e-05,
+      "loss": 1.5554,
+      "step": 18597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6588009595870972,
+      "learning_rate": 4.5086439333862016e-05,
+      "loss": 1.26,
+      "step": 18598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.527801752090454,
+      "learning_rate": 4.508617499339149e-05,
+      "loss": 1.2656,
+      "step": 18599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2127819061279297,
+      "learning_rate": 4.5085910652920964e-05,
+      "loss": 1.4485,
+      "step": 18600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.797106146812439,
+      "learning_rate": 4.508564631245044e-05,
+      "loss": 1.6725,
+      "step": 18601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3967416286468506,
+      "learning_rate": 4.508538197197991e-05,
+      "loss": 1.3399,
+      "step": 18602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5419163703918457,
+      "learning_rate": 4.508511763150939e-05,
+      "loss": 1.4704,
+      "step": 18603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3504295349121094,
+      "learning_rate": 4.508485329103886e-05,
+      "loss": 1.5909,
+      "step": 18604
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7428075075149536,
+      "learning_rate": 4.508458895056834e-05,
+      "loss": 1.4331,
+      "step": 18605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8461467027664185,
+      "learning_rate": 4.5084324610097805e-05,
+      "loss": 1.312,
+      "step": 18606
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3987468481063843,
+      "learning_rate": 4.508406026962728e-05,
+      "loss": 1.2069,
+      "step": 18607
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.972194790840149,
+      "learning_rate": 4.508379592915676e-05,
+      "loss": 1.5614,
+      "step": 18608
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4517782926559448,
+      "learning_rate": 4.5083531588686226e-05,
+      "loss": 1.3502,
+      "step": 18609
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8906910419464111,
+      "learning_rate": 4.5083267248215706e-05,
+      "loss": 1.3762,
+      "step": 18610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.301780939102173,
+      "learning_rate": 4.508300290774517e-05,
+      "loss": 1.5026,
+      "step": 18611
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.151930570602417,
+      "learning_rate": 4.508273856727465e-05,
+      "loss": 1.4399,
+      "step": 18612
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0775163173675537,
+      "learning_rate": 4.508247422680413e-05,
+      "loss": 1.4566,
+      "step": 18613
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5995804071426392,
+      "learning_rate": 4.50822098863336e-05,
+      "loss": 1.3346,
+      "step": 18614
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.072665214538574,
+      "learning_rate": 4.5081945545863074e-05,
+      "loss": 1.5298,
+      "step": 18615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8710191249847412,
+      "learning_rate": 4.508168120539255e-05,
+      "loss": 1.3374,
+      "step": 18616
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5895919799804688,
+      "learning_rate": 4.508141686492202e-05,
+      "loss": 1.4658,
+      "step": 18617
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8574610948562622,
+      "learning_rate": 4.5081152524451495e-05,
+      "loss": 1.4348,
+      "step": 18618
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6502835750579834,
+      "learning_rate": 4.508088818398097e-05,
+      "loss": 1.3353,
+      "step": 18619
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8463302850723267,
+      "learning_rate": 4.508062384351044e-05,
+      "loss": 2.5856,
+      "step": 18620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8599083423614502,
+      "learning_rate": 4.5080359503039915e-05,
+      "loss": 1.4622,
+      "step": 18621
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.683795213699341,
+      "learning_rate": 4.508009516256939e-05,
+      "loss": 1.5809,
+      "step": 18622
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4165986776351929,
+      "learning_rate": 4.507983082209887e-05,
+      "loss": 1.2007,
+      "step": 18623
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8772077560424805,
+      "learning_rate": 4.5079566481628336e-05,
+      "loss": 1.417,
+      "step": 18624
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.357613205909729,
+      "learning_rate": 4.5079302141157816e-05,
+      "loss": 1.2871,
+      "step": 18625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7015615701675415,
+      "learning_rate": 4.507903780068728e-05,
+      "loss": 1.3221,
+      "step": 18626
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0597572326660156,
+      "learning_rate": 4.5078773460216764e-05,
+      "loss": 1.4812,
+      "step": 18627
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.901268720626831,
+      "learning_rate": 4.507850911974624e-05,
+      "loss": 1.2563,
+      "step": 18628
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6996128559112549,
+      "learning_rate": 4.507824477927571e-05,
+      "loss": 1.2673,
+      "step": 18629
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3092539310455322,
+      "learning_rate": 4.5077980438805184e-05,
+      "loss": 1.5744,
+      "step": 18630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.351083993911743,
+      "learning_rate": 4.507771609833466e-05,
+      "loss": 3.1166,
+      "step": 18631
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.032844066619873,
+      "learning_rate": 4.507745175786413e-05,
+      "loss": 1.5805,
+      "step": 18632
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.757100820541382,
+      "learning_rate": 4.5077187417393605e-05,
+      "loss": 1.4855,
+      "step": 18633
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.909074068069458,
+      "learning_rate": 4.507692307692308e-05,
+      "loss": 1.4261,
+      "step": 18634
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9846839904785156,
+      "learning_rate": 4.507665873645255e-05,
+      "loss": 1.4022,
+      "step": 18635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8616582155227661,
+      "learning_rate": 4.5076394395982026e-05,
+      "loss": 1.3778,
+      "step": 18636
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.958202600479126,
+      "learning_rate": 4.50761300555115e-05,
+      "loss": 1.3435,
+      "step": 18637
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8175137042999268,
+      "learning_rate": 4.507586571504098e-05,
+      "loss": 1.296,
+      "step": 18638
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6810791492462158,
+      "learning_rate": 4.5075601374570446e-05,
+      "loss": 1.3017,
+      "step": 18639
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7541571855545044,
+      "learning_rate": 4.507533703409993e-05,
+      "loss": 1.6333,
+      "step": 18640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0220749378204346,
+      "learning_rate": 4.5075072693629393e-05,
+      "loss": 1.3243,
+      "step": 18641
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6297950744628906,
+      "learning_rate": 4.5074808353158874e-05,
+      "loss": 1.1858,
+      "step": 18642
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4211716651916504,
+      "learning_rate": 4.507454401268835e-05,
+      "loss": 1.3307,
+      "step": 18643
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0661683082580566,
+      "learning_rate": 4.5074279672217814e-05,
+      "loss": 1.4896,
+      "step": 18644
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.198638916015625,
+      "learning_rate": 4.5074015331747295e-05,
+      "loss": 1.4082,
+      "step": 18645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4202344417572021,
+      "learning_rate": 4.507375099127676e-05,
+      "loss": 1.3412,
+      "step": 18646
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.790000557899475,
+      "learning_rate": 4.507348665080624e-05,
+      "loss": 1.3962,
+      "step": 18647
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.705125570297241,
+      "learning_rate": 4.5073222310335715e-05,
+      "loss": 1.566,
+      "step": 18648
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.248668909072876,
+      "learning_rate": 4.507295796986519e-05,
+      "loss": 1.3839,
+      "step": 18649
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8077936172485352,
+      "learning_rate": 4.507269362939466e-05,
+      "loss": 1.2608,
+      "step": 18650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9062273502349854,
+      "learning_rate": 4.5072429288924136e-05,
+      "loss": 1.5148,
+      "step": 18651
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7708673477172852,
+      "learning_rate": 4.507216494845361e-05,
+      "loss": 1.3083,
+      "step": 18652
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.009734869003296,
+      "learning_rate": 4.507190060798308e-05,
+      "loss": 1.3531,
+      "step": 18653
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.883080005645752,
+      "learning_rate": 4.507163626751256e-05,
+      "loss": 1.3347,
+      "step": 18654
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.193650245666504,
+      "learning_rate": 4.507137192704203e-05,
+      "loss": 1.4497,
+      "step": 18655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0920238494873047,
+      "learning_rate": 4.5071107586571504e-05,
+      "loss": 1.284,
+      "step": 18656
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5301936864852905,
+      "learning_rate": 4.507084324610098e-05,
+      "loss": 1.2678,
+      "step": 18657
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.718352198600769,
+      "learning_rate": 4.507057890563046e-05,
+      "loss": 1.312,
+      "step": 18658
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0340187549591064,
+      "learning_rate": 4.5070314565159925e-05,
+      "loss": 1.4395,
+      "step": 18659
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9588489532470703,
+      "learning_rate": 4.5070050224689405e-05,
+      "loss": 1.3904,
+      "step": 18660
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5891295671463013,
+      "learning_rate": 4.506978588421887e-05,
+      "loss": 1.3269,
+      "step": 18661
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1757622957229614,
+      "learning_rate": 4.506952154374835e-05,
+      "loss": 1.5434,
+      "step": 18662
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.560701608657837,
+      "learning_rate": 4.5069257203277826e-05,
+      "loss": 1.4201,
+      "step": 18663
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6957112550735474,
+      "learning_rate": 4.50689928628073e-05,
+      "loss": 1.4317,
+      "step": 18664
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6363630294799805,
+      "learning_rate": 4.506872852233677e-05,
+      "loss": 1.3326,
+      "step": 18665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3908783197402954,
+      "learning_rate": 4.5068464181866246e-05,
+      "loss": 1.8125,
+      "step": 18666
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7408041954040527,
+      "learning_rate": 4.506819984139572e-05,
+      "loss": 1.309,
+      "step": 18667
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4271748065948486,
+      "learning_rate": 4.5067935500925193e-05,
+      "loss": 1.2576,
+      "step": 18668
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0371153354644775,
+      "learning_rate": 4.506767116045467e-05,
+      "loss": 1.4947,
+      "step": 18669
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6363463401794434,
+      "learning_rate": 4.506740681998414e-05,
+      "loss": 1.324,
+      "step": 18670
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8224248886108398,
+      "learning_rate": 4.5067142479513614e-05,
+      "loss": 1.4127,
+      "step": 18671
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8995696306228638,
+      "learning_rate": 4.506687813904309e-05,
+      "loss": 1.4042,
+      "step": 18672
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3622446060180664,
+      "learning_rate": 4.506661379857257e-05,
+      "loss": 1.2682,
+      "step": 18673
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8129205703735352,
+      "learning_rate": 4.5066349458102035e-05,
+      "loss": 1.4337,
+      "step": 18674
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0104918479919434,
+      "learning_rate": 4.5066085117631515e-05,
+      "loss": 1.4901,
+      "step": 18675
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6660317182540894,
+      "learning_rate": 4.506582077716098e-05,
+      "loss": 1.3803,
+      "step": 18676
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2499654293060303,
+      "learning_rate": 4.506555643669046e-05,
+      "loss": 1.4799,
+      "step": 18677
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.405377984046936,
+      "learning_rate": 4.5065292096219936e-05,
+      "loss": 1.3199,
+      "step": 18678
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7735059261322021,
+      "learning_rate": 4.50650277557494e-05,
+      "loss": 1.0431,
+      "step": 18679
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7532079219818115,
+      "learning_rate": 4.506476341527888e-05,
+      "loss": 1.4039,
+      "step": 18680
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7166038751602173,
+      "learning_rate": 4.506449907480835e-05,
+      "loss": 1.3136,
+      "step": 18681
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7511050701141357,
+      "learning_rate": 4.506423473433783e-05,
+      "loss": 1.4621,
+      "step": 18682
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3977328538894653,
+      "learning_rate": 4.5063970393867304e-05,
+      "loss": 1.3085,
+      "step": 18683
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2829339504241943,
+      "learning_rate": 4.506370605339678e-05,
+      "loss": 1.4611,
+      "step": 18684
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6394736766815186,
+      "learning_rate": 4.506344171292625e-05,
+      "loss": 1.6082,
+      "step": 18685
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6396806240081787,
+      "learning_rate": 4.5063177372455725e-05,
+      "loss": 1.2592,
+      "step": 18686
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8154144287109375,
+      "learning_rate": 4.50629130319852e-05,
+      "loss": 1.4811,
+      "step": 18687
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.280428409576416,
+      "learning_rate": 4.506264869151468e-05,
+      "loss": 1.3639,
+      "step": 18688
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2543234825134277,
+      "learning_rate": 4.5062384351044145e-05,
+      "loss": 1.5013,
+      "step": 18689
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9018373489379883,
+      "learning_rate": 4.506212001057362e-05,
+      "loss": 1.5741,
+      "step": 18690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9405279159545898,
+      "learning_rate": 4.506185567010309e-05,
+      "loss": 1.529,
+      "step": 18691
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8544549942016602,
+      "learning_rate": 4.5061591329632566e-05,
+      "loss": 1.2623,
+      "step": 18692
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6165595054626465,
+      "learning_rate": 4.5061326989162046e-05,
+      "loss": 1.4397,
+      "step": 18693
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6775516271591187,
+      "learning_rate": 4.506106264869151e-05,
+      "loss": 1.5267,
+      "step": 18694
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8939471244812012,
+      "learning_rate": 4.5060798308220993e-05,
+      "loss": 1.5354,
+      "step": 18695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7719792127609253,
+      "learning_rate": 4.506053396775046e-05,
+      "loss": 1.4036,
+      "step": 18696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9495141506195068,
+      "learning_rate": 4.506026962727994e-05,
+      "loss": 1.3842,
+      "step": 18697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8301968574523926,
+      "learning_rate": 4.5060005286809414e-05,
+      "loss": 1.256,
+      "step": 18698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5066076517105103,
+      "learning_rate": 4.505974094633889e-05,
+      "loss": 1.3193,
+      "step": 18699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7338340282440186,
+      "learning_rate": 4.505947660586836e-05,
+      "loss": 1.5608,
+      "step": 18700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5826812982559204,
+      "learning_rate": 4.5059212265397835e-05,
+      "loss": 1.1754,
+      "step": 18701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.000697135925293,
+      "learning_rate": 4.505894792492731e-05,
+      "loss": 1.21,
+      "step": 18702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.896748661994934,
+      "learning_rate": 4.505868358445678e-05,
+      "loss": 1.4587,
+      "step": 18703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6516801118850708,
+      "learning_rate": 4.5058419243986256e-05,
+      "loss": 1.375,
+      "step": 18704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7164757251739502,
+      "learning_rate": 4.505815490351573e-05,
+      "loss": 1.3546,
+      "step": 18705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1835758686065674,
+      "learning_rate": 4.50578905630452e-05,
+      "loss": 1.7744,
+      "step": 18706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7189033031463623,
+      "learning_rate": 4.5057626222574676e-05,
+      "loss": 1.1441,
+      "step": 18707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0448379516601562,
+      "learning_rate": 4.505736188210416e-05,
+      "loss": 1.3703,
+      "step": 18708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.581630825996399,
+      "learning_rate": 4.5057097541633623e-05,
+      "loss": 1.3583,
+      "step": 18709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.375360369682312,
+      "learning_rate": 4.5056833201163104e-05,
+      "loss": 1.3211,
+      "step": 18710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9011569023132324,
+      "learning_rate": 4.505656886069257e-05,
+      "loss": 1.3553,
+      "step": 18711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6777740716934204,
+      "learning_rate": 4.505630452022205e-05,
+      "loss": 1.2443,
+      "step": 18712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.98274564743042,
+      "learning_rate": 4.5056040179751524e-05,
+      "loss": 1.4414,
+      "step": 18713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9244745969772339,
+      "learning_rate": 4.5055775839281e-05,
+      "loss": 1.4747,
+      "step": 18714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3550018072128296,
+      "learning_rate": 4.505551149881047e-05,
+      "loss": 2.0838,
+      "step": 18715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.552113652229309,
+      "learning_rate": 4.505524715833994e-05,
+      "loss": 1.3375,
+      "step": 18716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7437005043029785,
+      "learning_rate": 4.505498281786942e-05,
+      "loss": 1.4105,
+      "step": 18717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.357898235321045,
+      "learning_rate": 4.505471847739889e-05,
+      "loss": 1.5593,
+      "step": 18718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4918525218963623,
+      "learning_rate": 4.5054454136928366e-05,
+      "loss": 1.2143,
+      "step": 18719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4880496263504028,
+      "learning_rate": 4.505418979645784e-05,
+      "loss": 1.2643,
+      "step": 18720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.51100492477417,
+      "learning_rate": 4.505392545598731e-05,
+      "loss": 1.5419,
+      "step": 18721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8602831363677979,
+      "learning_rate": 4.505366111551679e-05,
+      "loss": 1.3653,
+      "step": 18722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8766263723373413,
+      "learning_rate": 4.505339677504627e-05,
+      "loss": 1.4706,
+      "step": 18723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.566667914390564,
+      "learning_rate": 4.5053132434575734e-05,
+      "loss": 1.2296,
+      "step": 18724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.857430338859558,
+      "learning_rate": 4.505286809410521e-05,
+      "loss": 1.3562,
+      "step": 18725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0818986892700195,
+      "learning_rate": 4.505260375363468e-05,
+      "loss": 1.4435,
+      "step": 18726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.472353458404541,
+      "learning_rate": 4.5052339413164154e-05,
+      "loss": 1.3526,
+      "step": 18727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6142075061798096,
+      "learning_rate": 4.5052075072693635e-05,
+      "loss": 1.3575,
+      "step": 18728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.942373275756836,
+      "learning_rate": 4.50518107322231e-05,
+      "loss": 1.6003,
+      "step": 18729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.780434012413025,
+      "learning_rate": 4.505154639175258e-05,
+      "loss": 1.3093,
+      "step": 18730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7657921314239502,
+      "learning_rate": 4.505128205128205e-05,
+      "loss": 1.319,
+      "step": 18731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.762046456336975,
+      "learning_rate": 4.505101771081153e-05,
+      "loss": 1.4186,
+      "step": 18732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8101993799209595,
+      "learning_rate": 4.5050753370341e-05,
+      "loss": 1.3257,
+      "step": 18733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.377429485321045,
+      "learning_rate": 4.5050489029870476e-05,
+      "loss": 1.5303,
+      "step": 18734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7176405191421509,
+      "learning_rate": 4.505022468939995e-05,
+      "loss": 1.2913,
+      "step": 18735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4565207958221436,
+      "learning_rate": 4.5049960348929423e-05,
+      "loss": 1.4706,
+      "step": 18736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3676037788391113,
+      "learning_rate": 4.50496960084589e-05,
+      "loss": 1.2356,
+      "step": 18737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.132882595062256,
+      "learning_rate": 4.504943166798837e-05,
+      "loss": 1.4509,
+      "step": 18738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6431248188018799,
+      "learning_rate": 4.5049167327517844e-05,
+      "loss": 1.3206,
+      "step": 18739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.825270414352417,
+      "learning_rate": 4.504890298704732e-05,
+      "loss": 1.3801,
+      "step": 18740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1988120079040527,
+      "learning_rate": 4.504863864657679e-05,
+      "loss": 1.3215,
+      "step": 18741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6851541996002197,
+      "learning_rate": 4.5048374306106265e-05,
+      "loss": 1.1979,
+      "step": 18742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6154781579971313,
+      "learning_rate": 4.5048109965635745e-05,
+      "loss": 1.3159,
+      "step": 18743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.709588646888733,
+      "learning_rate": 4.504784562516521e-05,
+      "loss": 1.4108,
+      "step": 18744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2546348571777344,
+      "learning_rate": 4.504758128469469e-05,
+      "loss": 1.5699,
+      "step": 18745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0429909229278564,
+      "learning_rate": 4.504731694422416e-05,
+      "loss": 1.4952,
+      "step": 18746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2468953132629395,
+      "learning_rate": 4.504705260375364e-05,
+      "loss": 1.3313,
+      "step": 18747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2855606079101562,
+      "learning_rate": 4.504678826328311e-05,
+      "loss": 1.4337,
+      "step": 18748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.440698266029358,
+      "learning_rate": 4.5046523922812587e-05,
+      "loss": 2.4854,
+      "step": 18749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6151001453399658,
+      "learning_rate": 4.504625958234206e-05,
+      "loss": 1.4518,
+      "step": 18750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1541693210601807,
+      "learning_rate": 4.504599524187153e-05,
+      "loss": 1.4443,
+      "step": 18751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9432756900787354,
+      "learning_rate": 4.504573090140101e-05,
+      "loss": 1.496,
+      "step": 18752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8848240375518799,
+      "learning_rate": 4.504546656093048e-05,
+      "loss": 1.3332,
+      "step": 18753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7040305137634277,
+      "learning_rate": 4.5045202220459954e-05,
+      "loss": 1.3451,
+      "step": 18754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6546131372451782,
+      "learning_rate": 4.504493787998943e-05,
+      "loss": 1.1964,
+      "step": 18755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7916089296340942,
+      "learning_rate": 4.50446735395189e-05,
+      "loss": 1.2893,
+      "step": 18756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0891754627227783,
+      "learning_rate": 4.5044409199048375e-05,
+      "loss": 1.4411,
+      "step": 18757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8607220649719238,
+      "learning_rate": 4.5044144858577856e-05,
+      "loss": 1.5116,
+      "step": 18758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.929264545440674,
+      "learning_rate": 4.504388051810732e-05,
+      "loss": 1.8196,
+      "step": 18759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.666488528251648,
+      "learning_rate": 4.50436161776368e-05,
+      "loss": 1.3603,
+      "step": 18760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.223113775253296,
+      "learning_rate": 4.504335183716627e-05,
+      "loss": 1.4596,
+      "step": 18761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.646543264389038,
+      "learning_rate": 4.504308749669574e-05,
+      "loss": 1.3804,
+      "step": 18762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9011976718902588,
+      "learning_rate": 4.504282315622522e-05,
+      "loss": 1.3355,
+      "step": 18763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2753626108169556,
+      "learning_rate": 4.504255881575469e-05,
+      "loss": 2.0374,
+      "step": 18764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2566287517547607,
+      "learning_rate": 4.504229447528417e-05,
+      "loss": 1.4107,
+      "step": 18765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5257997512817383,
+      "learning_rate": 4.504203013481364e-05,
+      "loss": 2.5822,
+      "step": 18766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7901803255081177,
+      "learning_rate": 4.504176579434312e-05,
+      "loss": 1.2762,
+      "step": 18767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0518057346343994,
+      "learning_rate": 4.504150145387259e-05,
+      "loss": 1.606,
+      "step": 18768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.589530110359192,
+      "learning_rate": 4.5041237113402065e-05,
+      "loss": 1.3781,
+      "step": 18769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.802925944328308,
+      "learning_rate": 4.504097277293154e-05,
+      "loss": 1.4718,
+      "step": 18770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7052592039108276,
+      "learning_rate": 4.504070843246101e-05,
+      "loss": 1.2331,
+      "step": 18771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3671822547912598,
+      "learning_rate": 4.5040444091990485e-05,
+      "loss": 1.2828,
+      "step": 18772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.669143557548523,
+      "learning_rate": 4.504017975151996e-05,
+      "loss": 1.4422,
+      "step": 18773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1305301189422607,
+      "learning_rate": 4.503991541104943e-05,
+      "loss": 1.5185,
+      "step": 18774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.290645718574524,
+      "learning_rate": 4.5039651070578906e-05,
+      "loss": 1.3432,
+      "step": 18775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9305052757263184,
+      "learning_rate": 4.503938673010838e-05,
+      "loss": 1.4736,
+      "step": 18776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6019258499145508,
+      "learning_rate": 4.503912238963785e-05,
+      "loss": 1.3754,
+      "step": 18777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4204548597335815,
+      "learning_rate": 4.5038858049167334e-05,
+      "loss": 1.193,
+      "step": 18778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6873384714126587,
+      "learning_rate": 4.50385937086968e-05,
+      "loss": 1.4498,
+      "step": 18779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.08564829826355,
+      "learning_rate": 4.503832936822628e-05,
+      "loss": 1.2167,
+      "step": 18780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.058107614517212,
+      "learning_rate": 4.503806502775575e-05,
+      "loss": 1.4805,
+      "step": 18781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6796129941940308,
+      "learning_rate": 4.503780068728523e-05,
+      "loss": 1.4744,
+      "step": 18782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7048075199127197,
+      "learning_rate": 4.50375363468147e-05,
+      "loss": 1.3845,
+      "step": 18783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8031885623931885,
+      "learning_rate": 4.5037272006344175e-05,
+      "loss": 1.4422,
+      "step": 18784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.635045051574707,
+      "learning_rate": 4.503700766587365e-05,
+      "loss": 1.3852,
+      "step": 18785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5673383474349976,
+      "learning_rate": 4.503674332540312e-05,
+      "loss": 1.3973,
+      "step": 18786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.273575782775879,
+      "learning_rate": 4.5036478984932596e-05,
+      "loss": 1.5055,
+      "step": 18787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8780481815338135,
+      "learning_rate": 4.503621464446207e-05,
+      "loss": 1.5206,
+      "step": 18788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8389021158218384,
+      "learning_rate": 4.503595030399154e-05,
+      "loss": 1.3098,
+      "step": 18789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6281708478927612,
+      "learning_rate": 4.5035685963521017e-05,
+      "loss": 1.3743,
+      "step": 18790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0934383869171143,
+      "learning_rate": 4.503542162305049e-05,
+      "loss": 1.5718,
+      "step": 18791
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7673149108886719,
+      "learning_rate": 4.5035157282579964e-05,
+      "loss": 1.5194,
+      "step": 18792
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.694891333580017,
+      "learning_rate": 4.5034892942109444e-05,
+      "loss": 1.3154,
+      "step": 18793
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4900332689285278,
+      "learning_rate": 4.503462860163891e-05,
+      "loss": 1.2242,
+      "step": 18794
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6032638549804688,
+      "learning_rate": 4.503436426116839e-05,
+      "loss": 1.414,
+      "step": 18795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7144242525100708,
+      "learning_rate": 4.503409992069786e-05,
+      "loss": 1.2868,
+      "step": 18796
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.726707100868225,
+      "learning_rate": 4.503383558022733e-05,
+      "loss": 1.363,
+      "step": 18797
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.668465495109558,
+      "learning_rate": 4.503357123975681e-05,
+      "loss": 1.3748,
+      "step": 18798
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.442283868789673,
+      "learning_rate": 4.503330689928628e-05,
+      "loss": 1.5011,
+      "step": 18799
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7983161211013794,
+      "learning_rate": 4.503304255881576e-05,
+      "loss": 1.513,
+      "step": 18800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.044313430786133,
+      "learning_rate": 4.5032778218345226e-05,
+      "loss": 1.4281,
+      "step": 18801
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9209731817245483,
+      "learning_rate": 4.5032513877874706e-05,
+      "loss": 1.3617,
+      "step": 18802
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9508137702941895,
+      "learning_rate": 4.503224953740418e-05,
+      "loss": 1.3912,
+      "step": 18803
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5691263675689697,
+      "learning_rate": 4.503198519693365e-05,
+      "loss": 1.3451,
+      "step": 18804
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.444373607635498,
+      "learning_rate": 4.503172085646313e-05,
+      "loss": 1.2451,
+      "step": 18805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.492837905883789,
+      "learning_rate": 4.50314565159926e-05,
+      "loss": 1.3185,
+      "step": 18806
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.927902102470398,
+      "learning_rate": 4.5031192175522074e-05,
+      "loss": 1.3777,
+      "step": 18807
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.119035482406616,
+      "learning_rate": 4.503092783505155e-05,
+      "loss": 1.419,
+      "step": 18808
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5169912576675415,
+      "learning_rate": 4.503066349458102e-05,
+      "loss": 1.4041,
+      "step": 18809
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4066245555877686,
+      "learning_rate": 4.5030399154110495e-05,
+      "loss": 1.2026,
+      "step": 18810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.66216242313385,
+      "learning_rate": 4.503013481363997e-05,
+      "loss": 1.4036,
+      "step": 18811
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3893945217132568,
+      "learning_rate": 4.502987047316944e-05,
+      "loss": 1.2172,
+      "step": 18812
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.83411705493927,
+      "learning_rate": 4.502960613269892e-05,
+      "loss": 1.3798,
+      "step": 18813
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.820953607559204,
+      "learning_rate": 4.502934179222839e-05,
+      "loss": 1.3952,
+      "step": 18814
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1845455169677734,
+      "learning_rate": 4.502907745175787e-05,
+      "loss": 1.3537,
+      "step": 18815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7999242544174194,
+      "learning_rate": 4.5028813111287336e-05,
+      "loss": 1.413,
+      "step": 18816
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.892518162727356,
+      "learning_rate": 4.5028548770816817e-05,
+      "loss": 1.5711,
+      "step": 18817
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4730734825134277,
+      "learning_rate": 4.502828443034629e-05,
+      "loss": 1.5623,
+      "step": 18818
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3528120517730713,
+      "learning_rate": 4.5028020089875764e-05,
+      "loss": 1.3389,
+      "step": 18819
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.060446262359619,
+      "learning_rate": 4.502775574940524e-05,
+      "loss": 1.6056,
+      "step": 18820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9158610105514526,
+      "learning_rate": 4.502749140893471e-05,
+      "loss": 2.1822,
+      "step": 18821
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.506056308746338,
+      "learning_rate": 4.5027227068464184e-05,
+      "loss": 1.6357,
+      "step": 18822
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.724593162536621,
+      "learning_rate": 4.502696272799366e-05,
+      "loss": 1.4345,
+      "step": 18823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5153263807296753,
+      "learning_rate": 4.502669838752313e-05,
+      "loss": 1.3602,
+      "step": 18824
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9610952138900757,
+      "learning_rate": 4.5026434047052605e-05,
+      "loss": 1.3698,
+      "step": 18825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8778414726257324,
+      "learning_rate": 4.502616970658208e-05,
+      "loss": 1.4932,
+      "step": 18826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.797865629196167,
+      "learning_rate": 4.502590536611155e-05,
+      "loss": 1.3844,
+      "step": 18827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5640031099319458,
+      "learning_rate": 4.502564102564103e-05,
+      "loss": 1.2986,
+      "step": 18828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5723836421966553,
+      "learning_rate": 4.50253766851705e-05,
+      "loss": 1.3607,
+      "step": 18829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9035762548446655,
+      "learning_rate": 4.502511234469998e-05,
+      "loss": 1.2337,
+      "step": 18830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5119378566741943,
+      "learning_rate": 4.5024848004229447e-05,
+      "loss": 1.407,
+      "step": 18831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8585178852081299,
+      "learning_rate": 4.502458366375893e-05,
+      "loss": 1.3621,
+      "step": 18832
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4312214851379395,
+      "learning_rate": 4.50243193232884e-05,
+      "loss": 1.4343,
+      "step": 18833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.052435874938965,
+      "learning_rate": 4.502405498281787e-05,
+      "loss": 2.4937,
+      "step": 18834
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8336738348007202,
+      "learning_rate": 4.502379064234735e-05,
+      "loss": 1.2868,
+      "step": 18835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.023853063583374,
+      "learning_rate": 4.5023526301876814e-05,
+      "loss": 1.5076,
+      "step": 18836
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1061339378356934,
+      "learning_rate": 4.5023261961406295e-05,
+      "loss": 1.5137,
+      "step": 18837
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.200347661972046,
+      "learning_rate": 4.502299762093577e-05,
+      "loss": 1.5141,
+      "step": 18838
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.465723991394043,
+      "learning_rate": 4.502273328046524e-05,
+      "loss": 1.5016,
+      "step": 18839
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4332950115203857,
+      "learning_rate": 4.5022468939994715e-05,
+      "loss": 1.2386,
+      "step": 18840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5203222036361694,
+      "learning_rate": 4.502220459952419e-05,
+      "loss": 1.3795,
+      "step": 18841
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.154205322265625,
+      "learning_rate": 4.502194025905366e-05,
+      "loss": 1.3915,
+      "step": 18842
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5884058475494385,
+      "learning_rate": 4.5021675918583136e-05,
+      "loss": 1.3637,
+      "step": 18843
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3709906339645386,
+      "learning_rate": 4.502141157811261e-05,
+      "loss": 1.0909,
+      "step": 18844
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5772045850753784,
+      "learning_rate": 4.502114723764208e-05,
+      "loss": 1.2536,
+      "step": 18845
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1211097240448,
+      "learning_rate": 4.502088289717156e-05,
+      "loss": 1.3868,
+      "step": 18846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4376848936080933,
+      "learning_rate": 4.502061855670103e-05,
+      "loss": 1.2677,
+      "step": 18847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9527807235717773,
+      "learning_rate": 4.502035421623051e-05,
+      "loss": 1.475,
+      "step": 18848
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7614805698394775,
+      "learning_rate": 4.502008987575998e-05,
+      "loss": 1.5863,
+      "step": 18849
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.489134430885315,
+      "learning_rate": 4.501982553528946e-05,
+      "loss": 1.1739,
+      "step": 18850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1744749546051025,
+      "learning_rate": 4.5019561194818925e-05,
+      "loss": 1.4656,
+      "step": 18851
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4703288078308105,
+      "learning_rate": 4.5019296854348405e-05,
+      "loss": 1.959,
+      "step": 18852
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5817562341690063,
+      "learning_rate": 4.501903251387788e-05,
+      "loss": 1.834,
+      "step": 18853
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.514019727706909,
+      "learning_rate": 4.501876817340735e-05,
+      "loss": 1.5418,
+      "step": 18854
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5161669254302979,
+      "learning_rate": 4.5018503832936826e-05,
+      "loss": 1.2968,
+      "step": 18855
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.903137445449829,
+      "learning_rate": 4.50182394924663e-05,
+      "loss": 1.536,
+      "step": 18856
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6147297620773315,
+      "learning_rate": 4.501797515199577e-05,
+      "loss": 1.2756,
+      "step": 18857
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.655632495880127,
+      "learning_rate": 4.5017710811525246e-05,
+      "loss": 1.4339,
+      "step": 18858
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3735779523849487,
+      "learning_rate": 4.501744647105472e-05,
+      "loss": 1.2148,
+      "step": 18859
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.806778073310852,
+      "learning_rate": 4.5017182130584194e-05,
+      "loss": 1.3761,
+      "step": 18860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6361968517303467,
+      "learning_rate": 4.501691779011367e-05,
+      "loss": 1.3564,
+      "step": 18861
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.771554946899414,
+      "learning_rate": 4.501665344964314e-05,
+      "loss": 1.324,
+      "step": 18862
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0391271114349365,
+      "learning_rate": 4.501638910917262e-05,
+      "loss": 1.4941,
+      "step": 18863
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3589295148849487,
+      "learning_rate": 4.501612476870209e-05,
+      "loss": 1.2253,
+      "step": 18864
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8331425189971924,
+      "learning_rate": 4.501586042823157e-05,
+      "loss": 1.4688,
+      "step": 18865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.711896538734436,
+      "learning_rate": 4.5015596087761035e-05,
+      "loss": 1.2371,
+      "step": 18866
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.655598521232605,
+      "learning_rate": 4.5015331747290515e-05,
+      "loss": 1.2678,
+      "step": 18867
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9826602935791016,
+      "learning_rate": 4.501506740681999e-05,
+      "loss": 1.3281,
+      "step": 18868
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0972740650177,
+      "learning_rate": 4.5014803066349456e-05,
+      "loss": 1.3747,
+      "step": 18869
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.694362759590149,
+      "learning_rate": 4.5014538725878936e-05,
+      "loss": 1.2403,
+      "step": 18870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.803987741470337,
+      "learning_rate": 4.50142743854084e-05,
+      "loss": 1.44,
+      "step": 18871
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.710888147354126,
+      "learning_rate": 4.501401004493788e-05,
+      "loss": 1.3561,
+      "step": 18872
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9550548791885376,
+      "learning_rate": 4.501374570446736e-05,
+      "loss": 1.3744,
+      "step": 18873
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.50192129611969,
+      "learning_rate": 4.501348136399683e-05,
+      "loss": 1.2981,
+      "step": 18874
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.498536229133606,
+      "learning_rate": 4.5013217023526304e-05,
+      "loss": 1.3901,
+      "step": 18875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.847664713859558,
+      "learning_rate": 4.501295268305578e-05,
+      "loss": 1.1945,
+      "step": 18876
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1040897369384766,
+      "learning_rate": 4.501268834258525e-05,
+      "loss": 1.4745,
+      "step": 18877
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6325714588165283,
+      "learning_rate": 4.501242400211473e-05,
+      "loss": 1.4672,
+      "step": 18878
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5320141315460205,
+      "learning_rate": 4.50121596616442e-05,
+      "loss": 1.3277,
+      "step": 18879
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7878056764602661,
+      "learning_rate": 4.501189532117367e-05,
+      "loss": 1.2828,
+      "step": 18880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2182273864746094,
+      "learning_rate": 4.5011630980703145e-05,
+      "loss": 1.46,
+      "step": 18881
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0789809226989746,
+      "learning_rate": 4.501136664023262e-05,
+      "loss": 1.3854,
+      "step": 18882
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6519529819488525,
+      "learning_rate": 4.50111022997621e-05,
+      "loss": 1.4627,
+      "step": 18883
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8756059408187866,
+      "learning_rate": 4.5010837959291566e-05,
+      "loss": 1.4861,
+      "step": 18884
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.061689853668213,
+      "learning_rate": 4.5010573618821046e-05,
+      "loss": 1.504,
+      "step": 18885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.551749348640442,
+      "learning_rate": 4.501030927835051e-05,
+      "loss": 1.3341,
+      "step": 18886
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7037938833236694,
+      "learning_rate": 4.5010044937879994e-05,
+      "loss": 1.4217,
+      "step": 18887
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9249160289764404,
+      "learning_rate": 4.500978059740947e-05,
+      "loss": 1.4584,
+      "step": 18888
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.777479648590088,
+      "learning_rate": 4.500951625693894e-05,
+      "loss": 1.3614,
+      "step": 18889
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5308492183685303,
+      "learning_rate": 4.5009251916468414e-05,
+      "loss": 1.526,
+      "step": 18890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.461800217628479,
+      "learning_rate": 4.500898757599789e-05,
+      "loss": 1.2646,
+      "step": 18891
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4517337083816528,
+      "learning_rate": 4.500872323552736e-05,
+      "loss": 1.3768,
+      "step": 18892
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0398130416870117,
+      "learning_rate": 4.5008458895056835e-05,
+      "loss": 1.4224,
+      "step": 18893
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7190078496932983,
+      "learning_rate": 4.500819455458631e-05,
+      "loss": 1.3871,
+      "step": 18894
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.868730902671814,
+      "learning_rate": 4.500793021411578e-05,
+      "loss": 1.3498,
+      "step": 18895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6163568496704102,
+      "learning_rate": 4.5007665873645256e-05,
+      "loss": 1.3567,
+      "step": 18896
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7291913032531738,
+      "learning_rate": 4.500740153317473e-05,
+      "loss": 1.774,
+      "step": 18897
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7548311948776245,
+      "learning_rate": 4.500713719270421e-05,
+      "loss": 1.3527,
+      "step": 18898
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.547333002090454,
+      "learning_rate": 4.5006872852233676e-05,
+      "loss": 1.3196,
+      "step": 18899
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6515333652496338,
+      "learning_rate": 4.500660851176316e-05,
+      "loss": 1.3543,
+      "step": 18900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6169875860214233,
+      "learning_rate": 4.5006344171292624e-05,
+      "loss": 1.2671,
+      "step": 18901
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6954666376113892,
+      "learning_rate": 4.5006079830822104e-05,
+      "loss": 1.3176,
+      "step": 18902
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4463399648666382,
+      "learning_rate": 4.500581549035158e-05,
+      "loss": 1.2299,
+      "step": 18903
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2690670490264893,
+      "learning_rate": 4.500555114988105e-05,
+      "loss": 1.429,
+      "step": 18904
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8700834512710571,
+      "learning_rate": 4.5005286809410525e-05,
+      "loss": 1.3407,
+      "step": 18905
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9279834032058716,
+      "learning_rate": 4.500502246893999e-05,
+      "loss": 1.3961,
+      "step": 18906
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.002092123031616,
+      "learning_rate": 4.500475812846947e-05,
+      "loss": 1.2903,
+      "step": 18907
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.921968936920166,
+      "learning_rate": 4.5004493787998945e-05,
+      "loss": 1.1276,
+      "step": 18908
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.860427737236023,
+      "learning_rate": 4.500422944752842e-05,
+      "loss": 1.4831,
+      "step": 18909
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.186617612838745,
+      "learning_rate": 4.500396510705789e-05,
+      "loss": 1.3955,
+      "step": 18910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7190563678741455,
+      "learning_rate": 4.5003700766587366e-05,
+      "loss": 1.326,
+      "step": 18911
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9717907905578613,
+      "learning_rate": 4.500343642611684e-05,
+      "loss": 2.2236,
+      "step": 18912
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3388917446136475,
+      "learning_rate": 4.500317208564632e-05,
+      "loss": 2.3535,
+      "step": 18913
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5390079021453857,
+      "learning_rate": 4.500290774517579e-05,
+      "loss": 1.6005,
+      "step": 18914
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.229340076446533,
+      "learning_rate": 4.500264340470526e-05,
+      "loss": 1.619,
+      "step": 18915
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.050342559814453,
+      "learning_rate": 4.5002379064234734e-05,
+      "loss": 1.4588,
+      "step": 18916
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.469076156616211,
+      "learning_rate": 4.500211472376421e-05,
+      "loss": 1.6179,
+      "step": 18917
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4714274406433105,
+      "learning_rate": 4.500185038329369e-05,
+      "loss": 1.3346,
+      "step": 18918
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7202918529510498,
+      "learning_rate": 4.5001586042823155e-05,
+      "loss": 1.2529,
+      "step": 18919
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.665218472480774,
+      "learning_rate": 4.5001321702352635e-05,
+      "loss": 1.3167,
+      "step": 18920
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6187387704849243,
+      "learning_rate": 4.50010573618821e-05,
+      "loss": 1.3873,
+      "step": 18921
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7261536121368408,
+      "learning_rate": 4.500079302141158e-05,
+      "loss": 1.3778,
+      "step": 18922
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.476129174232483,
+      "learning_rate": 4.5000528680941056e-05,
+      "loss": 2.118,
+      "step": 18923
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6038349866867065,
+      "learning_rate": 4.500026434047053e-05,
+      "loss": 1.2055,
+      "step": 18924
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5639853477478027,
+      "learning_rate": 4.5e-05,
+      "loss": 1.9006,
+      "step": 18925
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4173394441604614,
+      "learning_rate": 4.4999735659529476e-05,
+      "loss": 1.2699,
+      "step": 18926
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.288437843322754,
+      "learning_rate": 4.499947131905895e-05,
+      "loss": 1.3381,
+      "step": 18927
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5468116998672485,
+      "learning_rate": 4.4999206978588424e-05,
+      "loss": 1.3924,
+      "step": 18928
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.437183141708374,
+      "learning_rate": 4.49989426381179e-05,
+      "loss": 1.323,
+      "step": 18929
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.971470832824707,
+      "learning_rate": 4.499867829764737e-05,
+      "loss": 1.3306,
+      "step": 18930
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5481141805648804,
+      "learning_rate": 4.4998413957176844e-05,
+      "loss": 1.161,
+      "step": 18931
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7911490201950073,
+      "learning_rate": 4.499814961670632e-05,
+      "loss": 1.3651,
+      "step": 18932
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.747546911239624,
+      "learning_rate": 4.49978852762358e-05,
+      "loss": 1.3663,
+      "step": 18933
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.014573335647583,
+      "learning_rate": 4.4997620935765265e-05,
+      "loss": 1.318,
+      "step": 18934
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.752362847328186,
+      "learning_rate": 4.4997356595294745e-05,
+      "loss": 1.2396,
+      "step": 18935
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.926221251487732,
+      "learning_rate": 4.499709225482421e-05,
+      "loss": 1.2423,
+      "step": 18936
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7574316263198853,
+      "learning_rate": 4.499682791435369e-05,
+      "loss": 1.3743,
+      "step": 18937
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6288385391235352,
+      "learning_rate": 4.4996563573883166e-05,
+      "loss": 1.3993,
+      "step": 18938
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7316863536834717,
+      "learning_rate": 4.499629923341264e-05,
+      "loss": 1.3773,
+      "step": 18939
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3862358331680298,
+      "learning_rate": 4.499603489294211e-05,
+      "loss": 1.4775,
+      "step": 18940
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.843894362449646,
+      "learning_rate": 4.499577055247158e-05,
+      "loss": 1.4663,
+      "step": 18941
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.495229482650757,
+      "learning_rate": 4.499550621200106e-05,
+      "loss": 1.2827,
+      "step": 18942
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6667027473449707,
+      "learning_rate": 4.4995241871530534e-05,
+      "loss": 1.3364,
+      "step": 18943
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0159952640533447,
+      "learning_rate": 4.499497753106001e-05,
+      "loss": 1.1821,
+      "step": 18944
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3756392002105713,
+      "learning_rate": 4.499471319058948e-05,
+      "loss": 1.4451,
+      "step": 18945
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.76859450340271,
+      "learning_rate": 4.4994448850118955e-05,
+      "loss": 1.2813,
+      "step": 18946
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.103848695755005,
+      "learning_rate": 4.499418450964843e-05,
+      "loss": 1.44,
+      "step": 18947
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7588868141174316,
+      "learning_rate": 4.499392016917791e-05,
+      "loss": 1.3284,
+      "step": 18948
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7052501440048218,
+      "learning_rate": 4.4993655828707375e-05,
+      "loss": 1.1933,
+      "step": 18949
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6928415298461914,
+      "learning_rate": 4.4993391488236856e-05,
+      "loss": 1.3474,
+      "step": 18950
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0098049640655518,
+      "learning_rate": 4.499312714776632e-05,
+      "loss": 1.4013,
+      "step": 18951
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4893956184387207,
+      "learning_rate": 4.4992862807295796e-05,
+      "loss": 1.2075,
+      "step": 18952
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.38850474357605,
+      "learning_rate": 4.4992598466825276e-05,
+      "loss": 1.3388,
+      "step": 18953
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.782741665840149,
+      "learning_rate": 4.499233412635474e-05,
+      "loss": 1.249,
+      "step": 18954
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.638003945350647,
+      "learning_rate": 4.4992069785884224e-05,
+      "loss": 1.3357,
+      "step": 18955
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4004876613616943,
+      "learning_rate": 4.499180544541369e-05,
+      "loss": 1.4486,
+      "step": 18956
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0879554748535156,
+      "learning_rate": 4.499154110494317e-05,
+      "loss": 1.2539,
+      "step": 18957
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0561859607696533,
+      "learning_rate": 4.4991276764472644e-05,
+      "loss": 1.2425,
+      "step": 18958
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8445026874542236,
+      "learning_rate": 4.499101242400212e-05,
+      "loss": 1.2628,
+      "step": 18959
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0434958934783936,
+      "learning_rate": 4.499074808353159e-05,
+      "loss": 1.3401,
+      "step": 18960
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0875980854034424,
+      "learning_rate": 4.4990483743061065e-05,
+      "loss": 1.2259,
+      "step": 18961
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5090656280517578,
+      "learning_rate": 4.499021940259054e-05,
+      "loss": 1.1504,
+      "step": 18962
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.099785327911377,
+      "learning_rate": 4.498995506212001e-05,
+      "loss": 1.2838,
+      "step": 18963
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7631200551986694,
+      "learning_rate": 4.4989690721649486e-05,
+      "loss": 1.2603,
+      "step": 18964
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9331259727478027,
+      "learning_rate": 4.498942638117896e-05,
+      "loss": 1.2535,
+      "step": 18965
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.005549430847168,
+      "learning_rate": 4.498916204070843e-05,
+      "loss": 1.276,
+      "step": 18966
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9224525690078735,
+      "learning_rate": 4.4988897700237906e-05,
+      "loss": 1.1704,
+      "step": 18967
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7335336208343506,
+      "learning_rate": 4.498863335976739e-05,
+      "loss": 1.282,
+      "step": 18968
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.279285430908203,
+      "learning_rate": 4.4988369019296854e-05,
+      "loss": 1.4102,
+      "step": 18969
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.538515567779541,
+      "learning_rate": 4.4988104678826334e-05,
+      "loss": 1.4372,
+      "step": 18970
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0179288387298584,
+      "learning_rate": 4.49878403383558e-05,
+      "loss": 1.1552,
+      "step": 18971
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8480290174484253,
+      "learning_rate": 4.498757599788528e-05,
+      "loss": 1.2548,
+      "step": 18972
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.844973564147949,
+      "learning_rate": 4.4987311657414755e-05,
+      "loss": 1.4666,
+      "step": 18973
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0473945140838623,
+      "learning_rate": 4.498704731694423e-05,
+      "loss": 1.3429,
+      "step": 18974
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.448084831237793,
+      "learning_rate": 4.49867829764737e-05,
+      "loss": 1.6307,
+      "step": 18975
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.899462342262268,
+      "learning_rate": 4.4986518636003175e-05,
+      "loss": 1.3267,
+      "step": 18976
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9845894575119019,
+      "learning_rate": 4.498625429553265e-05,
+      "loss": 1.2768,
+      "step": 18977
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7540885210037231,
+      "learning_rate": 4.498598995506212e-05,
+      "loss": 1.2544,
+      "step": 18978
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9246407747268677,
+      "learning_rate": 4.4985725614591596e-05,
+      "loss": 1.2925,
+      "step": 18979
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9957995414733887,
+      "learning_rate": 4.498546127412107e-05,
+      "loss": 1.2653,
+      "step": 18980
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.702420949935913,
+      "learning_rate": 4.498519693365054e-05,
+      "loss": 1.2166,
+      "step": 18981
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.022162675857544,
+      "learning_rate": 4.498493259318002e-05,
+      "loss": 1.8102,
+      "step": 18982
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8466017246246338,
+      "learning_rate": 4.49846682527095e-05,
+      "loss": 1.308,
+      "step": 18983
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7952500581741333,
+      "learning_rate": 4.4984403912238964e-05,
+      "loss": 1.2527,
+      "step": 18984
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1484498977661133,
+      "learning_rate": 4.4984139571768444e-05,
+      "loss": 1.2788,
+      "step": 18985
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7220197916030884,
+      "learning_rate": 4.498387523129791e-05,
+      "loss": 1.2975,
+      "step": 18986
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.729022741317749,
+      "learning_rate": 4.4983610890827385e-05,
+      "loss": 1.2335,
+      "step": 18987
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6274667978286743,
+      "learning_rate": 4.4983346550356865e-05,
+      "loss": 1.2201,
+      "step": 18988
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5981248617172241,
+      "learning_rate": 4.498308220988633e-05,
+      "loss": 1.2537,
+      "step": 18989
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8233892917633057,
+      "learning_rate": 4.498281786941581e-05,
+      "loss": 1.3776,
+      "step": 18990
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.179053544998169,
+      "learning_rate": 4.498255352894528e-05,
+      "loss": 1.315,
+      "step": 18991
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.209291696548462,
+      "learning_rate": 4.498228918847476e-05,
+      "loss": 1.4218,
+      "step": 18992
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6795001029968262,
+      "learning_rate": 4.498202484800423e-05,
+      "loss": 1.043,
+      "step": 18993
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1564085483551025,
+      "learning_rate": 4.4981760507533706e-05,
+      "loss": 1.4472,
+      "step": 18994
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8590039014816284,
+      "learning_rate": 4.498149616706318e-05,
+      "loss": 1.323,
+      "step": 18995
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8639887571334839,
+      "learning_rate": 4.4981231826592653e-05,
+      "loss": 1.0564,
+      "step": 18996
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.624510645866394,
+      "learning_rate": 4.498096748612213e-05,
+      "loss": 1.1568,
+      "step": 18997
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.604355812072754,
+      "learning_rate": 4.49807031456516e-05,
+      "loss": 1.6088,
+      "step": 18998
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.026939868927002,
+      "learning_rate": 4.4980438805181074e-05,
+      "loss": 1.3018,
+      "step": 18999
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.738405466079712,
+      "learning_rate": 4.498017446471055e-05,
+      "loss": 1.244,
+      "step": 19000
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.50422739982605,
+      "learning_rate": 4.497991012424002e-05,
+      "loss": 1.4315,
+      "step": 19001
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8929296731948853,
+      "learning_rate": 4.4979645783769495e-05,
+      "loss": 1.3081,
+      "step": 19002
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3114044666290283,
+      "learning_rate": 4.4979381443298975e-05,
+      "loss": 1.4909,
+      "step": 19003
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2080495357513428,
+      "learning_rate": 4.497911710282844e-05,
+      "loss": 1.2306,
+      "step": 19004
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6855766773223877,
+      "learning_rate": 4.497885276235792e-05,
+      "loss": 1.2446,
+      "step": 19005
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2350475788116455,
+      "learning_rate": 4.497858842188739e-05,
+      "loss": 1.3757,
+      "step": 19006
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7979917526245117,
+      "learning_rate": 4.497832408141687e-05,
+      "loss": 1.2026,
+      "step": 19007
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.596797466278076,
+      "learning_rate": 4.497805974094634e-05,
+      "loss": 1.5984,
+      "step": 19008
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.642277717590332,
+      "learning_rate": 4.497779540047582e-05,
+      "loss": 1.1681,
+      "step": 19009
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.680129051208496,
+      "learning_rate": 4.497753106000529e-05,
+      "loss": 1.2331,
+      "step": 19010
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.823395848274231,
+      "learning_rate": 4.4977266719534764e-05,
+      "loss": 1.3265,
+      "step": 19011
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6443309783935547,
+      "learning_rate": 4.497700237906424e-05,
+      "loss": 1.2705,
+      "step": 19012
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.98727285861969,
+      "learning_rate": 4.497673803859371e-05,
+      "loss": 1.7756,
+      "step": 19013
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.255779266357422,
+      "learning_rate": 4.4976473698123185e-05,
+      "loss": 1.3636,
+      "step": 19014
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6305311918258667,
+      "learning_rate": 4.497620935765266e-05,
+      "loss": 1.1622,
+      "step": 19015
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1989519596099854,
+      "learning_rate": 4.497594501718213e-05,
+      "loss": 1.4393,
+      "step": 19016
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.757073998451233,
+      "learning_rate": 4.4975680676711605e-05,
+      "loss": 1.2578,
+      "step": 19017
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.795186996459961,
+      "learning_rate": 4.4975416336241086e-05,
+      "loss": 1.3064,
+      "step": 19018
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.96712064743042,
+      "learning_rate": 4.497515199577055e-05,
+      "loss": 1.3966,
+      "step": 19019
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9126343727111816,
+      "learning_rate": 4.497488765530003e-05,
+      "loss": 1.1256,
+      "step": 19020
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9692533016204834,
+      "learning_rate": 4.49746233148295e-05,
+      "loss": 1.1938,
+      "step": 19021
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.840267300605774,
+      "learning_rate": 4.497435897435898e-05,
+      "loss": 1.368,
+      "step": 19022
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9151357412338257,
+      "learning_rate": 4.4974094633888453e-05,
+      "loss": 1.3549,
+      "step": 19023
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8239878416061401,
+      "learning_rate": 4.497383029341792e-05,
+      "loss": 1.0634,
+      "step": 19024
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.080193281173706,
+      "learning_rate": 4.49735659529474e-05,
+      "loss": 1.2694,
+      "step": 19025
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0952420234680176,
+      "learning_rate": 4.497330161247687e-05,
+      "loss": 1.2373,
+      "step": 19026
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7875771522521973,
+      "learning_rate": 4.497303727200635e-05,
+      "loss": 1.2336,
+      "step": 19027
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.80414617061615,
+      "learning_rate": 4.497277293153582e-05,
+      "loss": 1.2228,
+      "step": 19028
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1226248741149902,
+      "learning_rate": 4.4972508591065295e-05,
+      "loss": 1.3845,
+      "step": 19029
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9821199178695679,
+      "learning_rate": 4.497224425059477e-05,
+      "loss": 1.2986,
+      "step": 19030
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.279230833053589,
+      "learning_rate": 4.497197991012424e-05,
+      "loss": 1.2902,
+      "step": 19031
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.5449296236038208,
+      "learning_rate": 4.4971715569653716e-05,
+      "loss": 1.1205,
+      "step": 19032
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6711735725402832,
+      "learning_rate": 4.497145122918319e-05,
+      "loss": 2.0459,
+      "step": 19033
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8054883480072021,
+      "learning_rate": 4.497118688871266e-05,
+      "loss": 1.3678,
+      "step": 19034
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6682944297790527,
+      "learning_rate": 4.4970922548242136e-05,
+      "loss": 1.3089,
+      "step": 19035
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.2290735244750977,
+      "learning_rate": 4.497065820777161e-05,
+      "loss": 1.1477,
+      "step": 19036
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6034367084503174,
+      "learning_rate": 4.4970393867301083e-05,
+      "loss": 1.2005,
+      "step": 19037
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.094357013702393,
+      "learning_rate": 4.4970129526830564e-05,
+      "loss": 1.2752,
+      "step": 19038
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.913709282875061,
+      "learning_rate": 4.496986518636003e-05,
+      "loss": 1.3057,
+      "step": 19039
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.968048334121704,
+      "learning_rate": 4.496960084588951e-05,
+      "loss": 1.3151,
+      "step": 19040
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7192158699035645,
+      "learning_rate": 4.496933650541898e-05,
+      "loss": 1.3632,
+      "step": 19041
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.909165620803833,
+      "learning_rate": 4.496907216494846e-05,
+      "loss": 1.1232,
+      "step": 19042
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7451112270355225,
+      "learning_rate": 4.496880782447793e-05,
+      "loss": 1.2609,
+      "step": 19043
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.077188014984131,
+      "learning_rate": 4.4968543484007405e-05,
+      "loss": 1.4356,
+      "step": 19044
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.4765913486480713,
+      "learning_rate": 4.496827914353688e-05,
+      "loss": 1.4417,
+      "step": 19045
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9295514822006226,
+      "learning_rate": 4.496801480306635e-05,
+      "loss": 1.1328,
+      "step": 19046
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8314605951309204,
+      "learning_rate": 4.4967750462595826e-05,
+      "loss": 1.2331,
+      "step": 19047
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.228410482406616,
+      "learning_rate": 4.49674861221253e-05,
+      "loss": 1.2555,
+      "step": 19048
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1871254444122314,
+      "learning_rate": 4.496722178165477e-05,
+      "loss": 1.1497,
+      "step": 19049
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.276749849319458,
+      "learning_rate": 4.496695744118425e-05,
+      "loss": 1.2722,
+      "step": 19050
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6776564121246338,
+      "learning_rate": 4.496669310071372e-05,
+      "loss": 1.311,
+      "step": 19051
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7983883619308472,
+      "learning_rate": 4.4966428760243194e-05,
+      "loss": 1.1683,
+      "step": 19052
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.862086296081543,
+      "learning_rate": 4.4966164419772674e-05,
+      "loss": 1.3589,
+      "step": 19053
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.078134059906006,
+      "learning_rate": 4.496590007930214e-05,
+      "loss": 1.2636,
+      "step": 19054
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7569347620010376,
+      "learning_rate": 4.496563573883162e-05,
+      "loss": 1.2874,
+      "step": 19055
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.540868878364563,
+      "learning_rate": 4.496537139836109e-05,
+      "loss": 1.2625,
+      "step": 19056
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8043147325515747,
+      "learning_rate": 4.496510705789057e-05,
+      "loss": 1.1187,
+      "step": 19057
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8764678239822388,
+      "learning_rate": 4.496484271742004e-05,
+      "loss": 1.3155,
+      "step": 19058
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7398502826690674,
+      "learning_rate": 4.496457837694951e-05,
+      "loss": 1.341,
+      "step": 19059
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0416908264160156,
+      "learning_rate": 4.496431403647899e-05,
+      "loss": 1.3092,
+      "step": 19060
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1776647567749023,
+      "learning_rate": 4.4964049696008456e-05,
+      "loss": 1.3884,
+      "step": 19061
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.334346055984497,
+      "learning_rate": 4.4963785355537936e-05,
+      "loss": 1.4035,
+      "step": 19062
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9420160055160522,
+      "learning_rate": 4.496352101506741e-05,
+      "loss": 1.414,
+      "step": 19063
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7324306964874268,
+      "learning_rate": 4.4963256674596883e-05,
+      "loss": 1.1088,
+      "step": 19064
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9616199731826782,
+      "learning_rate": 4.496299233412636e-05,
+      "loss": 1.2312,
+      "step": 19065
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1779532432556152,
+      "learning_rate": 4.496272799365583e-05,
+      "loss": 1.5303,
+      "step": 19066
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.5698843002319336,
+      "learning_rate": 4.4962463653185304e-05,
+      "loss": 1.1552,
+      "step": 19067
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.982722520828247,
+      "learning_rate": 4.4962199312714784e-05,
+      "loss": 1.3241,
+      "step": 19068
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8751839399337769,
+      "learning_rate": 4.496193497224425e-05,
+      "loss": 1.3331,
+      "step": 19069
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9199527502059937,
+      "learning_rate": 4.4961670631773725e-05,
+      "loss": 1.1218,
+      "step": 19070
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7597297430038452,
+      "learning_rate": 4.49614062913032e-05,
+      "loss": 1.3075,
+      "step": 19071
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7328872680664062,
+      "learning_rate": 4.496114195083267e-05,
+      "loss": 2.3935,
+      "step": 19072
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8705648183822632,
+      "learning_rate": 4.496087761036215e-05,
+      "loss": 1.3838,
+      "step": 19073
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.909209132194519,
+      "learning_rate": 4.496061326989162e-05,
+      "loss": 1.2723,
+      "step": 19074
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6942672729492188,
+      "learning_rate": 4.49603489294211e-05,
+      "loss": 1.2148,
+      "step": 19075
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6754388809204102,
+      "learning_rate": 4.4960084588950566e-05,
+      "loss": 1.2131,
+      "step": 19076
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.3111789226531982,
+      "learning_rate": 4.4959820248480047e-05,
+      "loss": 1.3213,
+      "step": 19077
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6148924827575684,
+      "learning_rate": 4.495955590800952e-05,
+      "loss": 1.308,
+      "step": 19078
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9588719606399536,
+      "learning_rate": 4.4959291567538994e-05,
+      "loss": 1.4116,
+      "step": 19079
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8912749290466309,
+      "learning_rate": 4.495902722706847e-05,
+      "loss": 1.2499,
+      "step": 19080
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0397586822509766,
+      "learning_rate": 4.495876288659794e-05,
+      "loss": 1.2774,
+      "step": 19081
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.657668948173523,
+      "learning_rate": 4.4958498546127414e-05,
+      "loss": 1.323,
+      "step": 19082
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.943359136581421,
+      "learning_rate": 4.495823420565689e-05,
+      "loss": 1.2955,
+      "step": 19083
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6201369762420654,
+      "learning_rate": 4.495796986518636e-05,
+      "loss": 1.2132,
+      "step": 19084
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8358229398727417,
+      "learning_rate": 4.4957705524715835e-05,
+      "loss": 1.2534,
+      "step": 19085
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8355220556259155,
+      "learning_rate": 4.495744118424531e-05,
+      "loss": 1.3142,
+      "step": 19086
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.330321788787842,
+      "learning_rate": 4.495717684377478e-05,
+      "loss": 1.3303,
+      "step": 19087
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6253126859664917,
+      "learning_rate": 4.495691250330426e-05,
+      "loss": 1.628,
+      "step": 19088
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7654695510864258,
+      "learning_rate": 4.495664816283373e-05,
+      "loss": 1.2828,
+      "step": 19089
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7312828302383423,
+      "learning_rate": 4.495638382236321e-05,
+      "loss": 1.2964,
+      "step": 19090
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7002589702606201,
+      "learning_rate": 4.4956119481892677e-05,
+      "loss": 1.1991,
+      "step": 19091
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9006391763687134,
+      "learning_rate": 4.495585514142216e-05,
+      "loss": 1.2678,
+      "step": 19092
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.883312463760376,
+      "learning_rate": 4.495559080095163e-05,
+      "loss": 1.3295,
+      "step": 19093
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9894309043884277,
+      "learning_rate": 4.4955326460481104e-05,
+      "loss": 1.3409,
+      "step": 19094
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9152590036392212,
+      "learning_rate": 4.495506212001058e-05,
+      "loss": 1.2166,
+      "step": 19095
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9294072389602661,
+      "learning_rate": 4.4954797779540044e-05,
+      "loss": 1.1695,
+      "step": 19096
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8420617580413818,
+      "learning_rate": 4.4954533439069525e-05,
+      "loss": 1.2372,
+      "step": 19097
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7471781969070435,
+      "learning_rate": 4.4954269098599e-05,
+      "loss": 1.1304,
+      "step": 19098
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8872325420379639,
+      "learning_rate": 4.495400475812847e-05,
+      "loss": 1.2307,
+      "step": 19099
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.050086259841919,
+      "learning_rate": 4.4953740417657946e-05,
+      "loss": 1.3501,
+      "step": 19100
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.760404109954834,
+      "learning_rate": 4.495347607718742e-05,
+      "loss": 1.5699,
+      "step": 19101
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.918548822402954,
+      "learning_rate": 4.495321173671689e-05,
+      "loss": 1.3402,
+      "step": 19102
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8520655632019043,
+      "learning_rate": 4.495294739624637e-05,
+      "loss": 1.3125,
+      "step": 19103
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.6005351543426514,
+      "learning_rate": 4.495268305577584e-05,
+      "loss": 1.4805,
+      "step": 19104
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7819446325302124,
+      "learning_rate": 4.495241871530531e-05,
+      "loss": 1.3085,
+      "step": 19105
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6133227348327637,
+      "learning_rate": 4.495215437483479e-05,
+      "loss": 1.2104,
+      "step": 19106
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7294198274612427,
+      "learning_rate": 4.495189003436426e-05,
+      "loss": 1.2215,
+      "step": 19107
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.3166885375976562,
+      "learning_rate": 4.495162569389374e-05,
+      "loss": 1.3516,
+      "step": 19108
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.145801544189453,
+      "learning_rate": 4.495136135342321e-05,
+      "loss": 1.3159,
+      "step": 19109
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.79939341545105,
+      "learning_rate": 4.495109701295269e-05,
+      "loss": 1.3265,
+      "step": 19110
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.876125454902649,
+      "learning_rate": 4.4950832672482155e-05,
+      "loss": 1.2261,
+      "step": 19111
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.153337240219116,
+      "learning_rate": 4.4950568332011635e-05,
+      "loss": 1.4258,
+      "step": 19112
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.3178153038024902,
+      "learning_rate": 4.495030399154111e-05,
+      "loss": 1.31,
+      "step": 19113
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.744149088859558,
+      "learning_rate": 4.495003965107058e-05,
+      "loss": 1.1348,
+      "step": 19114
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6453338861465454,
+      "learning_rate": 4.4949775310600056e-05,
+      "loss": 1.2709,
+      "step": 19115
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7982276678085327,
+      "learning_rate": 4.494951097012953e-05,
+      "loss": 1.2434,
+      "step": 19116
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9947562217712402,
+      "learning_rate": 4.4949246629659e-05,
+      "loss": 1.2668,
+      "step": 19117
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.680473804473877,
+      "learning_rate": 4.4948982289188477e-05,
+      "loss": 1.2689,
+      "step": 19118
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.2933082580566406,
+      "learning_rate": 4.494871794871795e-05,
+      "loss": 1.4237,
+      "step": 19119
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7717550992965698,
+      "learning_rate": 4.4948453608247424e-05,
+      "loss": 1.2796,
+      "step": 19120
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7548292875289917,
+      "learning_rate": 4.49481892677769e-05,
+      "loss": 1.306,
+      "step": 19121
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.7388720512390137,
+      "learning_rate": 4.494792492730637e-05,
+      "loss": 1.4359,
+      "step": 19122
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.073915719985962,
+      "learning_rate": 4.494766058683585e-05,
+      "loss": 1.2074,
+      "step": 19123
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8015519380569458,
+      "learning_rate": 4.494739624636532e-05,
+      "loss": 1.2052,
+      "step": 19124
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.25029468536377,
+      "learning_rate": 4.49471319058948e-05,
+      "loss": 1.3674,
+      "step": 19125
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.019298553466797,
+      "learning_rate": 4.4946867565424265e-05,
+      "loss": 1.2828,
+      "step": 19126
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.504011392593384,
+      "learning_rate": 4.4946603224953745e-05,
+      "loss": 1.4257,
+      "step": 19127
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8868296146392822,
+      "learning_rate": 4.494633888448322e-05,
+      "loss": 1.3555,
+      "step": 19128
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.039342164993286,
+      "learning_rate": 4.494607454401269e-05,
+      "loss": 1.4194,
+      "step": 19129
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.960120916366577,
+      "learning_rate": 4.4945810203542166e-05,
+      "loss": 2.1674,
+      "step": 19130
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.0417580604553223,
+      "learning_rate": 4.494554586307164e-05,
+      "loss": 1.2653,
+      "step": 19131
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.067103385925293,
+      "learning_rate": 4.494528152260111e-05,
+      "loss": 1.3726,
+      "step": 19132
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.165020227432251,
+      "learning_rate": 4.494501718213059e-05,
+      "loss": 1.3099,
+      "step": 19133
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8740004301071167,
+      "learning_rate": 4.494475284166006e-05,
+      "loss": 1.3612,
+      "step": 19134
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7105724811553955,
+      "learning_rate": 4.4944488501189534e-05,
+      "loss": 1.3255,
+      "step": 19135
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.962218165397644,
+      "learning_rate": 4.494422416071901e-05,
+      "loss": 1.5209,
+      "step": 19136
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.6405067443847656,
+      "learning_rate": 4.494395982024848e-05,
+      "loss": 1.2635,
+      "step": 19137
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.929821491241455,
+      "learning_rate": 4.494369547977796e-05,
+      "loss": 1.1753,
+      "step": 19138
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6670544147491455,
+      "learning_rate": 4.494343113930743e-05,
+      "loss": 1.1881,
+      "step": 19139
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.4877331256866455,
+      "learning_rate": 4.494316679883691e-05,
+      "loss": 1.1884,
+      "step": 19140
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.695108413696289,
+      "learning_rate": 4.4942902458366375e-05,
+      "loss": 1.128,
+      "step": 19141
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.553450345993042,
+      "learning_rate": 4.494263811789585e-05,
+      "loss": 1.5414,
+      "step": 19142
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8679442405700684,
+      "learning_rate": 4.494237377742533e-05,
+      "loss": 1.459,
+      "step": 19143
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7193658351898193,
+      "learning_rate": 4.4942109436954796e-05,
+      "loss": 1.2958,
+      "step": 19144
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.060476303100586,
+      "learning_rate": 4.4941845096484277e-05,
+      "loss": 1.2625,
+      "step": 19145
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0103061199188232,
+      "learning_rate": 4.494158075601374e-05,
+      "loss": 1.277,
+      "step": 19146
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0573055744171143,
+      "learning_rate": 4.4941316415543224e-05,
+      "loss": 1.3564,
+      "step": 19147
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8948172330856323,
+      "learning_rate": 4.49410520750727e-05,
+      "loss": 1.4519,
+      "step": 19148
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.5573229789733887,
+      "learning_rate": 4.494078773460217e-05,
+      "loss": 1.4617,
+      "step": 19149
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.4376554489135742,
+      "learning_rate": 4.4940523394131644e-05,
+      "loss": 1.0988,
+      "step": 19150
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.3165040016174316,
+      "learning_rate": 4.494025905366112e-05,
+      "loss": 1.3277,
+      "step": 19151
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.847992181777954,
+      "learning_rate": 4.493999471319059e-05,
+      "loss": 1.2169,
+      "step": 19152
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6361907720565796,
+      "learning_rate": 4.4939730372720065e-05,
+      "loss": 1.3625,
+      "step": 19153
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9746673107147217,
+      "learning_rate": 4.493946603224954e-05,
+      "loss": 1.3963,
+      "step": 19154
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7930313348770142,
+      "learning_rate": 4.493920169177901e-05,
+      "loss": 1.2383,
+      "step": 19155
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.841571569442749,
+      "learning_rate": 4.4938937351308486e-05,
+      "loss": 1.2594,
+      "step": 19156
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8562343120574951,
+      "learning_rate": 4.493867301083796e-05,
+      "loss": 1.2319,
+      "step": 19157
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9194018840789795,
+      "learning_rate": 4.493840867036744e-05,
+      "loss": 1.2417,
+      "step": 19158
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7903082370758057,
+      "learning_rate": 4.4938144329896907e-05,
+      "loss": 1.2128,
+      "step": 19159
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.5375638008117676,
+      "learning_rate": 4.493787998942639e-05,
+      "loss": 1.477,
+      "step": 19160
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8325655460357666,
+      "learning_rate": 4.4937615648955854e-05,
+      "loss": 1.2389,
+      "step": 19161
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9960370063781738,
+      "learning_rate": 4.4937351308485334e-05,
+      "loss": 1.3582,
+      "step": 19162
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7285990715026855,
+      "learning_rate": 4.493708696801481e-05,
+      "loss": 1.205,
+      "step": 19163
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.822647213935852,
+      "learning_rate": 4.493682262754428e-05,
+      "loss": 1.259,
+      "step": 19164
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0787384510040283,
+      "learning_rate": 4.4936558287073755e-05,
+      "loss": 1.3835,
+      "step": 19165
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.3457603454589844,
+      "learning_rate": 4.493629394660323e-05,
+      "loss": 1.2631,
+      "step": 19166
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.923194169998169,
+      "learning_rate": 4.49360296061327e-05,
+      "loss": 1.3809,
+      "step": 19167
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6653060913085938,
+      "learning_rate": 4.4935765265662175e-05,
+      "loss": 1.3283,
+      "step": 19168
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0350418090820312,
+      "learning_rate": 4.493550092519165e-05,
+      "loss": 1.3531,
+      "step": 19169
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.4486308097839355,
+      "learning_rate": 4.493523658472112e-05,
+      "loss": 1.0496,
+      "step": 19170
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.986548662185669,
+      "learning_rate": 4.4934972244250596e-05,
+      "loss": 1.2884,
+      "step": 19171
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.68609881401062,
+      "learning_rate": 4.493470790378007e-05,
+      "loss": 1.3605,
+      "step": 19172
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.914050579071045,
+      "learning_rate": 4.493444356330955e-05,
+      "loss": 1.2637,
+      "step": 19173
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9613786935806274,
+      "learning_rate": 4.493417922283902e-05,
+      "loss": 1.2766,
+      "step": 19174
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.053525447845459,
+      "learning_rate": 4.49339148823685e-05,
+      "loss": 1.2002,
+      "step": 19175
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7567285299301147,
+      "learning_rate": 4.4933650541897964e-05,
+      "loss": 1.389,
+      "step": 19176
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7648029327392578,
+      "learning_rate": 4.493338620142744e-05,
+      "loss": 1.1908,
+      "step": 19177
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9709206819534302,
+      "learning_rate": 4.493312186095692e-05,
+      "loss": 1.3368,
+      "step": 19178
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.099256753921509,
+      "learning_rate": 4.4932857520486385e-05,
+      "loss": 1.2237,
+      "step": 19179
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0147266387939453,
+      "learning_rate": 4.4932593180015865e-05,
+      "loss": 1.0971,
+      "step": 19180
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0677990913391113,
+      "learning_rate": 4.493232883954533e-05,
+      "loss": 1.3052,
+      "step": 19181
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1974897384643555,
+      "learning_rate": 4.493206449907481e-05,
+      "loss": 1.4922,
+      "step": 19182
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8950968980789185,
+      "learning_rate": 4.4931800158604286e-05,
+      "loss": 1.2494,
+      "step": 19183
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.962710380554199,
+      "learning_rate": 4.493153581813376e-05,
+      "loss": 1.9009,
+      "step": 19184
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.2469794750213623,
+      "learning_rate": 4.493127147766323e-05,
+      "loss": 1.3324,
+      "step": 19185
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.2408735752105713,
+      "learning_rate": 4.4931007137192706e-05,
+      "loss": 1.3963,
+      "step": 19186
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8972076177597046,
+      "learning_rate": 4.493074279672218e-05,
+      "loss": 1.2824,
+      "step": 19187
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8632392883300781,
+      "learning_rate": 4.4930478456251654e-05,
+      "loss": 1.6544,
+      "step": 19188
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9285720586776733,
+      "learning_rate": 4.493021411578113e-05,
+      "loss": 1.317,
+      "step": 19189
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8160847425460815,
+      "learning_rate": 4.49299497753106e-05,
+      "loss": 2.0969,
+      "step": 19190
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.078167200088501,
+      "learning_rate": 4.4929685434840074e-05,
+      "loss": 1.3848,
+      "step": 19191
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9083002805709839,
+      "learning_rate": 4.492942109436955e-05,
+      "loss": 1.3943,
+      "step": 19192
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8066864013671875,
+      "learning_rate": 4.492915675389903e-05,
+      "loss": 1.2745,
+      "step": 19193
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.234915256500244,
+      "learning_rate": 4.4928892413428495e-05,
+      "loss": 1.3441,
+      "step": 19194
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.097276449203491,
+      "learning_rate": 4.4928628072957975e-05,
+      "loss": 1.4784,
+      "step": 19195
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9582124948501587,
+      "learning_rate": 4.492836373248744e-05,
+      "loss": 1.2314,
+      "step": 19196
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.516119956970215,
+      "learning_rate": 4.492809939201692e-05,
+      "loss": 1.325,
+      "step": 19197
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7255946397781372,
+      "learning_rate": 4.4927835051546396e-05,
+      "loss": 1.3029,
+      "step": 19198
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1058547496795654,
+      "learning_rate": 4.492757071107587e-05,
+      "loss": 1.4802,
+      "step": 19199
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.900527834892273,
+      "learning_rate": 4.492730637060534e-05,
+      "loss": 1.1967,
+      "step": 19200
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2436673641204834,
+      "learning_rate": 4.492704203013482e-05,
+      "loss": 1.2607,
+      "step": 19201
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0253803730010986,
+      "learning_rate": 4.492677768966429e-05,
+      "loss": 1.1834,
+      "step": 19202
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.0885472297668457,
+      "learning_rate": 4.4926513349193764e-05,
+      "loss": 1.3368,
+      "step": 19203
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.84451425075531,
+      "learning_rate": 4.492624900872324e-05,
+      "loss": 1.2774,
+      "step": 19204
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0265424251556396,
+      "learning_rate": 4.492598466825271e-05,
+      "loss": 1.4355,
+      "step": 19205
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8774042129516602,
+      "learning_rate": 4.4925720327782185e-05,
+      "loss": 1.2189,
+      "step": 19206
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.745908498764038,
+      "learning_rate": 4.492545598731166e-05,
+      "loss": 1.5588,
+      "step": 19207
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8704842329025269,
+      "learning_rate": 4.492519164684114e-05,
+      "loss": 1.346,
+      "step": 19208
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.860823631286621,
+      "learning_rate": 4.4924927306370605e-05,
+      "loss": 1.2602,
+      "step": 19209
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3528401851654053,
+      "learning_rate": 4.4924662965900086e-05,
+      "loss": 1.2945,
+      "step": 19210
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.891869306564331,
+      "learning_rate": 4.492439862542955e-05,
+      "loss": 1.3716,
+      "step": 19211
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0259995460510254,
+      "learning_rate": 4.492413428495903e-05,
+      "loss": 1.2997,
+      "step": 19212
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6570366621017456,
+      "learning_rate": 4.4923869944488506e-05,
+      "loss": 1.1833,
+      "step": 19213
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1175527572631836,
+      "learning_rate": 4.492360560401797e-05,
+      "loss": 1.4152,
+      "step": 19214
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6407830715179443,
+      "learning_rate": 4.4923341263547454e-05,
+      "loss": 1.4151,
+      "step": 19215
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9359033107757568,
+      "learning_rate": 4.492307692307692e-05,
+      "loss": 1.2791,
+      "step": 19216
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.462834119796753,
+      "learning_rate": 4.49228125826064e-05,
+      "loss": 1.4931,
+      "step": 19217
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1391026973724365,
+      "learning_rate": 4.4922548242135874e-05,
+      "loss": 1.4525,
+      "step": 19218
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.5905524492263794,
+      "learning_rate": 4.492228390166535e-05,
+      "loss": 1.173,
+      "step": 19219
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2409467697143555,
+      "learning_rate": 4.492201956119482e-05,
+      "loss": 1.4135,
+      "step": 19220
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.5819920301437378,
+      "learning_rate": 4.4921755220724295e-05,
+      "loss": 1.1283,
+      "step": 19221
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8527873754501343,
+      "learning_rate": 4.492149088025377e-05,
+      "loss": 1.3756,
+      "step": 19222
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0874147415161133,
+      "learning_rate": 4.492122653978324e-05,
+      "loss": 1.2337,
+      "step": 19223
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7419734001159668,
+      "learning_rate": 4.4920962199312716e-05,
+      "loss": 1.0991,
+      "step": 19224
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2306342124938965,
+      "learning_rate": 4.492069785884219e-05,
+      "loss": 1.285,
+      "step": 19225
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.596078872680664,
+      "learning_rate": 4.492043351837166e-05,
+      "loss": 1.3672,
+      "step": 19226
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8756513595581055,
+      "learning_rate": 4.4920169177901136e-05,
+      "loss": 1.3417,
+      "step": 19227
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6442930698394775,
+      "learning_rate": 4.491990483743062e-05,
+      "loss": 1.5013,
+      "step": 19228
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.8466291427612305,
+      "learning_rate": 4.4919640496960084e-05,
+      "loss": 1.108,
+      "step": 19229
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1100339889526367,
+      "learning_rate": 4.4919376156489564e-05,
+      "loss": 1.1727,
+      "step": 19230
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8042627573013306,
+      "learning_rate": 4.491911181601903e-05,
+      "loss": 1.2496,
+      "step": 19231
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7140165567398071,
+      "learning_rate": 4.491884747554851e-05,
+      "loss": 1.2002,
+      "step": 19232
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7098629474639893,
+      "learning_rate": 4.4918583135077985e-05,
+      "loss": 1.1408,
+      "step": 19233
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0085785388946533,
+      "learning_rate": 4.491831879460746e-05,
+      "loss": 1.3931,
+      "step": 19234
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8213616609573364,
+      "learning_rate": 4.491805445413693e-05,
+      "loss": 1.2122,
+      "step": 19235
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9926341772079468,
+      "learning_rate": 4.4917790113666405e-05,
+      "loss": 1.0276,
+      "step": 19236
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7578186988830566,
+      "learning_rate": 4.491752577319588e-05,
+      "loss": 1.6666,
+      "step": 19237
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8929526805877686,
+      "learning_rate": 4.491726143272535e-05,
+      "loss": 1.3609,
+      "step": 19238
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9367802143096924,
+      "learning_rate": 4.4916997092254826e-05,
+      "loss": 1.2793,
+      "step": 19239
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.157478094100952,
+      "learning_rate": 4.49167327517843e-05,
+      "loss": 1.3746,
+      "step": 19240
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9280400276184082,
+      "learning_rate": 4.491646841131377e-05,
+      "loss": 1.3582,
+      "step": 19241
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8991179466247559,
+      "learning_rate": 4.491620407084325e-05,
+      "loss": 1.2786,
+      "step": 19242
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.413191318511963,
+      "learning_rate": 4.491593973037273e-05,
+      "loss": 1.4057,
+      "step": 19243
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9756439924240112,
+      "learning_rate": 4.4915675389902194e-05,
+      "loss": 1.3121,
+      "step": 19244
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7530170679092407,
+      "learning_rate": 4.4915411049431674e-05,
+      "loss": 1.1494,
+      "step": 19245
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.743476390838623,
+      "learning_rate": 4.491514670896114e-05,
+      "loss": 1.3119,
+      "step": 19246
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8162027597427368,
+      "learning_rate": 4.491488236849062e-05,
+      "loss": 1.3022,
+      "step": 19247
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.143244743347168,
+      "learning_rate": 4.4914618028020095e-05,
+      "loss": 1.0258,
+      "step": 19248
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7401049137115479,
+      "learning_rate": 4.491435368754957e-05,
+      "loss": 1.315,
+      "step": 19249
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7519843578338623,
+      "learning_rate": 4.491408934707904e-05,
+      "loss": 1.0294,
+      "step": 19250
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8680747747421265,
+      "learning_rate": 4.491382500660851e-05,
+      "loss": 1.2521,
+      "step": 19251
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0132110118865967,
+      "learning_rate": 4.491356066613799e-05,
+      "loss": 1.3077,
+      "step": 19252
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0743215084075928,
+      "learning_rate": 4.491329632566746e-05,
+      "loss": 1.3985,
+      "step": 19253
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.7901082038879395,
+      "learning_rate": 4.4913031985196936e-05,
+      "loss": 1.354,
+      "step": 19254
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.303248882293701,
+      "learning_rate": 4.491276764472641e-05,
+      "loss": 1.2705,
+      "step": 19255
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7240495681762695,
+      "learning_rate": 4.4912503304255884e-05,
+      "loss": 1.2392,
+      "step": 19256
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.049536943435669,
+      "learning_rate": 4.491223896378536e-05,
+      "loss": 1.3578,
+      "step": 19257
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3407113552093506,
+      "learning_rate": 4.491197462331484e-05,
+      "loss": 1.3466,
+      "step": 19258
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1290550231933594,
+      "learning_rate": 4.4911710282844304e-05,
+      "loss": 1.2311,
+      "step": 19259
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7669458389282227,
+      "learning_rate": 4.491144594237378e-05,
+      "loss": 1.3401,
+      "step": 19260
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.010639190673828,
+      "learning_rate": 4.491118160190325e-05,
+      "loss": 2.2001,
+      "step": 19261
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.388758420944214,
+      "learning_rate": 4.4910917261432725e-05,
+      "loss": 1.4067,
+      "step": 19262
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2866480350494385,
+      "learning_rate": 4.4910652920962205e-05,
+      "loss": 1.4463,
+      "step": 19263
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.691046953201294,
+      "learning_rate": 4.491038858049167e-05,
+      "loss": 1.4199,
+      "step": 19264
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.057293176651001,
+      "learning_rate": 4.491012424002115e-05,
+      "loss": 1.2575,
+      "step": 19265
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9682307243347168,
+      "learning_rate": 4.490985989955062e-05,
+      "loss": 1.2555,
+      "step": 19266
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.315223455429077,
+      "learning_rate": 4.49095955590801e-05,
+      "loss": 1.3257,
+      "step": 19267
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8860437870025635,
+      "learning_rate": 4.490933121860957e-05,
+      "loss": 1.213,
+      "step": 19268
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9853026866912842,
+      "learning_rate": 4.490906687813905e-05,
+      "loss": 2.0795,
+      "step": 19269
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.44805645942688,
+      "learning_rate": 4.490880253766852e-05,
+      "loss": 1.2069,
+      "step": 19270
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8912553787231445,
+      "learning_rate": 4.4908538197197994e-05,
+      "loss": 1.226,
+      "step": 19271
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.2641489505767822,
+      "learning_rate": 4.490827385672747e-05,
+      "loss": 1.3151,
+      "step": 19272
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0758354663848877,
+      "learning_rate": 4.490800951625694e-05,
+      "loss": 1.3736,
+      "step": 19273
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.222510814666748,
+      "learning_rate": 4.4907745175786415e-05,
+      "loss": 1.3977,
+      "step": 19274
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7792795896530151,
+      "learning_rate": 4.490748083531589e-05,
+      "loss": 1.2805,
+      "step": 19275
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6855347156524658,
+      "learning_rate": 4.490721649484536e-05,
+      "loss": 1.1716,
+      "step": 19276
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7756139039993286,
+      "learning_rate": 4.4906952154374835e-05,
+      "loss": 1.2425,
+      "step": 19277
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.219165802001953,
+      "learning_rate": 4.4906687813904316e-05,
+      "loss": 1.3876,
+      "step": 19278
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2377521991729736,
+      "learning_rate": 4.490642347343378e-05,
+      "loss": 1.3904,
+      "step": 19279
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.019575834274292,
+      "learning_rate": 4.490615913296326e-05,
+      "loss": 1.1021,
+      "step": 19280
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8197669982910156,
+      "learning_rate": 4.490589479249273e-05,
+      "loss": 1.2729,
+      "step": 19281
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6433782577514648,
+      "learning_rate": 4.490563045202221e-05,
+      "loss": 1.886,
+      "step": 19282
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.285581588745117,
+      "learning_rate": 4.4905366111551684e-05,
+      "loss": 1.1763,
+      "step": 19283
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9584394693374634,
+      "learning_rate": 4.490510177108116e-05,
+      "loss": 1.1812,
+      "step": 19284
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9580739736557007,
+      "learning_rate": 4.490483743061063e-05,
+      "loss": 1.3581,
+      "step": 19285
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7861841917037964,
+      "learning_rate": 4.49045730901401e-05,
+      "loss": 1.1557,
+      "step": 19286
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7762023210525513,
+      "learning_rate": 4.490430874966958e-05,
+      "loss": 1.5306,
+      "step": 19287
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9069851636886597,
+      "learning_rate": 4.490404440919905e-05,
+      "loss": 1.1772,
+      "step": 19288
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.635970115661621,
+      "learning_rate": 4.4903780068728525e-05,
+      "loss": 1.2253,
+      "step": 19289
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.701803207397461,
+      "learning_rate": 4.4903515728258e-05,
+      "loss": 1.4744,
+      "step": 19290
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.0370237827301025,
+      "learning_rate": 4.490325138778747e-05,
+      "loss": 1.5678,
+      "step": 19291
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6363118886947632,
+      "learning_rate": 4.4902987047316946e-05,
+      "loss": 1.2327,
+      "step": 19292
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1428046226501465,
+      "learning_rate": 4.4902722706846426e-05,
+      "loss": 1.399,
+      "step": 19293
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.982653260231018,
+      "learning_rate": 4.490245836637589e-05,
+      "loss": 1.2889,
+      "step": 19294
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6814454793930054,
+      "learning_rate": 4.4902194025905366e-05,
+      "loss": 1.2674,
+      "step": 19295
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9995205402374268,
+      "learning_rate": 4.490192968543484e-05,
+      "loss": 1.2538,
+      "step": 19296
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.970499038696289,
+      "learning_rate": 4.4901665344964314e-05,
+      "loss": 1.3595,
+      "step": 19297
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.211965322494507,
+      "learning_rate": 4.4901401004493794e-05,
+      "loss": 1.8142,
+      "step": 19298
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8287242650985718,
+      "learning_rate": 4.490113666402326e-05,
+      "loss": 1.3657,
+      "step": 19299
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2207086086273193,
+      "learning_rate": 4.490087232355274e-05,
+      "loss": 1.8976,
+      "step": 19300
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.137557029724121,
+      "learning_rate": 4.490060798308221e-05,
+      "loss": 1.4744,
+      "step": 19301
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3219962120056152,
+      "learning_rate": 4.490034364261169e-05,
+      "loss": 1.304,
+      "step": 19302
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.781079888343811,
+      "learning_rate": 4.490007930214116e-05,
+      "loss": 1.1596,
+      "step": 19303
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8007848262786865,
+      "learning_rate": 4.4899814961670635e-05,
+      "loss": 1.2391,
+      "step": 19304
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.333777904510498,
+      "learning_rate": 4.489955062120011e-05,
+      "loss": 1.7168,
+      "step": 19305
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.844735860824585,
+      "learning_rate": 4.489928628072958e-05,
+      "loss": 1.2532,
+      "step": 19306
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3823277950286865,
+      "learning_rate": 4.4899021940259056e-05,
+      "loss": 1.4602,
+      "step": 19307
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7791359424591064,
+      "learning_rate": 4.489875759978853e-05,
+      "loss": 1.3419,
+      "step": 19308
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7760206460952759,
+      "learning_rate": 4.4898493259318e-05,
+      "loss": 1.1285,
+      "step": 19309
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8120990991592407,
+      "learning_rate": 4.489822891884748e-05,
+      "loss": 1.0833,
+      "step": 19310
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.966342568397522,
+      "learning_rate": 4.489796457837695e-05,
+      "loss": 1.796,
+      "step": 19311
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8935999870300293,
+      "learning_rate": 4.4897700237906424e-05,
+      "loss": 1.3382,
+      "step": 19312
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7265359163284302,
+      "learning_rate": 4.4897435897435904e-05,
+      "loss": 1.186,
+      "step": 19313
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.5000107288360596,
+      "learning_rate": 4.489717155696537e-05,
+      "loss": 1.373,
+      "step": 19314
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9547901153564453,
+      "learning_rate": 4.489690721649485e-05,
+      "loss": 1.2739,
+      "step": 19315
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7140395641326904,
+      "learning_rate": 4.489664287602432e-05,
+      "loss": 1.2593,
+      "step": 19316
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.793019413948059,
+      "learning_rate": 4.48963785355538e-05,
+      "loss": 1.9151,
+      "step": 19317
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8892441987991333,
+      "learning_rate": 4.489611419508327e-05,
+      "loss": 1.4059,
+      "step": 19318
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2362730503082275,
+      "learning_rate": 4.4895849854612746e-05,
+      "loss": 1.2518,
+      "step": 19319
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7819030284881592,
+      "learning_rate": 4.489558551414222e-05,
+      "loss": 1.6763,
+      "step": 19320
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.073113203048706,
+      "learning_rate": 4.489532117367169e-05,
+      "loss": 1.3249,
+      "step": 19321
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.701375961303711,
+      "learning_rate": 4.4895056833201166e-05,
+      "loss": 1.2243,
+      "step": 19322
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9571399688720703,
+      "learning_rate": 4.489479249273064e-05,
+      "loss": 1.2881,
+      "step": 19323
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6224496364593506,
+      "learning_rate": 4.4894528152260113e-05,
+      "loss": 1.4626,
+      "step": 19324
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0936760902404785,
+      "learning_rate": 4.489426381178959e-05,
+      "loss": 1.3006,
+      "step": 19325
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7841730117797852,
+      "learning_rate": 4.489399947131906e-05,
+      "loss": 1.3155,
+      "step": 19326
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2169086933135986,
+      "learning_rate": 4.4893735130848534e-05,
+      "loss": 1.4189,
+      "step": 19327
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8468257188796997,
+      "learning_rate": 4.4893470790378015e-05,
+      "loss": 1.1479,
+      "step": 19328
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.845907211303711,
+      "learning_rate": 4.489320644990748e-05,
+      "loss": 1.2563,
+      "step": 19329
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.168272018432617,
+      "learning_rate": 4.489294210943696e-05,
+      "loss": 1.294,
+      "step": 19330
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7851349115371704,
+      "learning_rate": 4.489267776896643e-05,
+      "loss": 1.2182,
+      "step": 19331
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8027076721191406,
+      "learning_rate": 4.48924134284959e-05,
+      "loss": 1.2535,
+      "step": 19332
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.01827073097229,
+      "learning_rate": 4.489214908802538e-05,
+      "loss": 1.2833,
+      "step": 19333
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.200185775756836,
+      "learning_rate": 4.489188474755485e-05,
+      "loss": 1.2441,
+      "step": 19334
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.46811580657959,
+      "learning_rate": 4.489162040708433e-05,
+      "loss": 1.3767,
+      "step": 19335
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9114625453948975,
+      "learning_rate": 4.4891356066613796e-05,
+      "loss": 1.3429,
+      "step": 19336
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3167388439178467,
+      "learning_rate": 4.489109172614328e-05,
+      "loss": 1.4541,
+      "step": 19337
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9185771942138672,
+      "learning_rate": 4.489082738567275e-05,
+      "loss": 1.2849,
+      "step": 19338
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2615773677825928,
+      "learning_rate": 4.4890563045202224e-05,
+      "loss": 1.3555,
+      "step": 19339
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6149516105651855,
+      "learning_rate": 4.48902987047317e-05,
+      "loss": 1.2426,
+      "step": 19340
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.4894678592681885,
+      "learning_rate": 4.489003436426117e-05,
+      "loss": 1.4002,
+      "step": 19341
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.80233633518219,
+      "learning_rate": 4.4889770023790645e-05,
+      "loss": 1.2561,
+      "step": 19342
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.529322624206543,
+      "learning_rate": 4.488950568332012e-05,
+      "loss": 1.2731,
+      "step": 19343
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.889374852180481,
+      "learning_rate": 4.488924134284959e-05,
+      "loss": 1.1923,
+      "step": 19344
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.4338492155075073,
+      "learning_rate": 4.4888977002379065e-05,
+      "loss": 1.8378,
+      "step": 19345
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0441081523895264,
+      "learning_rate": 4.488871266190854e-05,
+      "loss": 1.3525,
+      "step": 19346
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.692466378211975,
+      "learning_rate": 4.488844832143801e-05,
+      "loss": 1.2988,
+      "step": 19347
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6321804523468018,
+      "learning_rate": 4.488818398096749e-05,
+      "loss": 1.1203,
+      "step": 19348
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.129274845123291,
+      "learning_rate": 4.488791964049696e-05,
+      "loss": 1.2566,
+      "step": 19349
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.4033594131469727,
+      "learning_rate": 4.488765530002644e-05,
+      "loss": 1.3386,
+      "step": 19350
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0975842475891113,
+      "learning_rate": 4.488739095955591e-05,
+      "loss": 1.3623,
+      "step": 19351
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0986328125,
+      "learning_rate": 4.488712661908539e-05,
+      "loss": 1.2928,
+      "step": 19352
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.5869543552398682,
+      "learning_rate": 4.488686227861486e-05,
+      "loss": 1.2536,
+      "step": 19353
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3094139099121094,
+      "learning_rate": 4.4886597938144334e-05,
+      "loss": 1.2778,
+      "step": 19354
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.773836374282837,
+      "learning_rate": 4.488633359767381e-05,
+      "loss": 1.1839,
+      "step": 19355
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6440919637680054,
+      "learning_rate": 4.488606925720328e-05,
+      "loss": 1.0505,
+      "step": 19356
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.1055548191070557,
+      "learning_rate": 4.4885804916732755e-05,
+      "loss": 1.3715,
+      "step": 19357
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.5193328857421875,
+      "learning_rate": 4.488554057626223e-05,
+      "loss": 1.4145,
+      "step": 19358
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.039820671081543,
+      "learning_rate": 4.48852762357917e-05,
+      "loss": 1.4446,
+      "step": 19359
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.614546775817871,
+      "learning_rate": 4.4885011895321176e-05,
+      "loss": 1.2106,
+      "step": 19360
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1428704261779785,
+      "learning_rate": 4.488474755485065e-05,
+      "loss": 1.341,
+      "step": 19361
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6158561706542969,
+      "learning_rate": 4.488448321438012e-05,
+      "loss": 1.1709,
+      "step": 19362
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6271494626998901,
+      "learning_rate": 4.48842188739096e-05,
+      "loss": 1.2425,
+      "step": 19363
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.698266625404358,
+      "learning_rate": 4.488395453343907e-05,
+      "loss": 1.0293,
+      "step": 19364
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.195671796798706,
+      "learning_rate": 4.488369019296855e-05,
+      "loss": 1.3395,
+      "step": 19365
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.5792348384857178,
+      "learning_rate": 4.488342585249802e-05,
+      "loss": 1.1655,
+      "step": 19366
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.221224784851074,
+      "learning_rate": 4.488316151202749e-05,
+      "loss": 1.3641,
+      "step": 19367
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9094616174697876,
+      "learning_rate": 4.488289717155697e-05,
+      "loss": 1.3494,
+      "step": 19368
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.804124355316162,
+      "learning_rate": 4.488263283108644e-05,
+      "loss": 1.292,
+      "step": 19369
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8730541467666626,
+      "learning_rate": 4.488236849061592e-05,
+      "loss": 1.1527,
+      "step": 19370
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.612723708152771,
+      "learning_rate": 4.4882104150145385e-05,
+      "loss": 1.1276,
+      "step": 19371
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8422257900238037,
+      "learning_rate": 4.4881839809674865e-05,
+      "loss": 1.2871,
+      "step": 19372
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.5620014667510986,
+      "learning_rate": 4.488157546920434e-05,
+      "loss": 1.398,
+      "step": 19373
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1806514263153076,
+      "learning_rate": 4.488131112873381e-05,
+      "loss": 1.3756,
+      "step": 19374
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7129769325256348,
+      "learning_rate": 4.4881046788263286e-05,
+      "loss": 1.2436,
+      "step": 19375
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6327319145202637,
+      "learning_rate": 4.488078244779276e-05,
+      "loss": 0.9871,
+      "step": 19376
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.326815128326416,
+      "learning_rate": 4.488051810732223e-05,
+      "loss": 1.2505,
+      "step": 19377
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3888766765594482,
+      "learning_rate": 4.488025376685171e-05,
+      "loss": 1.7237,
+      "step": 19378
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8924002647399902,
+      "learning_rate": 4.487998942638118e-05,
+      "loss": 1.2389,
+      "step": 19379
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.838468313217163,
+      "learning_rate": 4.4879725085910654e-05,
+      "loss": 1.2171,
+      "step": 19380
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.026594400405884,
+      "learning_rate": 4.487946074544013e-05,
+      "loss": 2.01,
+      "step": 19381
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7461689710617065,
+      "learning_rate": 4.48791964049696e-05,
+      "loss": 1.3952,
+      "step": 19382
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.7401318550109863,
+      "learning_rate": 4.487893206449908e-05,
+      "loss": 1.2446,
+      "step": 19383
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8911786079406738,
+      "learning_rate": 4.487866772402855e-05,
+      "loss": 1.3055,
+      "step": 19384
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6821869611740112,
+      "learning_rate": 4.487840338355803e-05,
+      "loss": 1.168,
+      "step": 19385
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.038262367248535,
+      "learning_rate": 4.4878139043087495e-05,
+      "loss": 1.3425,
+      "step": 19386
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.2464380264282227,
+      "learning_rate": 4.4877874702616976e-05,
+      "loss": 1.3986,
+      "step": 19387
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9749395847320557,
+      "learning_rate": 4.487761036214645e-05,
+      "loss": 1.3118,
+      "step": 19388
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1000137329101562,
+      "learning_rate": 4.487734602167592e-05,
+      "loss": 1.3365,
+      "step": 19389
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.7809910774230957,
+      "learning_rate": 4.4877081681205396e-05,
+      "loss": 1.2702,
+      "step": 19390
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.074538230895996,
+      "learning_rate": 4.487681734073487e-05,
+      "loss": 1.8405,
+      "step": 19391
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9902985095977783,
+      "learning_rate": 4.4876553000264343e-05,
+      "loss": 1.3776,
+      "step": 19392
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.831910490989685,
+      "learning_rate": 4.487628865979382e-05,
+      "loss": 1.2193,
+      "step": 19393
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8536021709442139,
+      "learning_rate": 4.487602431932329e-05,
+      "loss": 1.2481,
+      "step": 19394
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.922465205192566,
+      "learning_rate": 4.4875759978852764e-05,
+      "loss": 1.2851,
+      "step": 19395
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.40779447555542,
+      "learning_rate": 4.487549563838224e-05,
+      "loss": 1.4062,
+      "step": 19396
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.041384696960449,
+      "learning_rate": 4.487523129791171e-05,
+      "loss": 1.3,
+      "step": 19397
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.874031662940979,
+      "learning_rate": 4.487496695744119e-05,
+      "loss": 1.2409,
+      "step": 19398
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9297326803207397,
+      "learning_rate": 4.487470261697066e-05,
+      "loss": 1.2671,
+      "step": 19399
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.4230129718780518,
+      "learning_rate": 4.487443827650014e-05,
+      "loss": 1.017,
+      "step": 19400
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.3601322174072266,
+      "learning_rate": 4.4874173936029606e-05,
+      "loss": 1.3294,
+      "step": 19401
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.4396297931671143,
+      "learning_rate": 4.4873909595559086e-05,
+      "loss": 1.2766,
+      "step": 19402
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9802087545394897,
+      "learning_rate": 4.487364525508856e-05,
+      "loss": 1.2523,
+      "step": 19403
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1941754817962646,
+      "learning_rate": 4.4873380914618026e-05,
+      "loss": 1.2211,
+      "step": 19404
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.5294172763824463,
+      "learning_rate": 4.4873116574147507e-05,
+      "loss": 1.2285,
+      "step": 19405
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5862094163894653,
+      "learning_rate": 4.4872852233676973e-05,
+      "loss": 1.2141,
+      "step": 19406
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1123030185699463,
+      "learning_rate": 4.4872587893206454e-05,
+      "loss": 1.2599,
+      "step": 19407
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5507861375808716,
+      "learning_rate": 4.487232355273593e-05,
+      "loss": 1.2522,
+      "step": 19408
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1608996391296387,
+      "learning_rate": 4.48720592122654e-05,
+      "loss": 1.3413,
+      "step": 19409
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8569588661193848,
+      "learning_rate": 4.4871794871794874e-05,
+      "loss": 1.2412,
+      "step": 19410
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.6944220066070557,
+      "learning_rate": 4.487153053132435e-05,
+      "loss": 1.2235,
+      "step": 19411
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.243606090545654,
+      "learning_rate": 4.487126619085382e-05,
+      "loss": 1.6231,
+      "step": 19412
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.457931399345398,
+      "learning_rate": 4.4871001850383295e-05,
+      "loss": 1.0716,
+      "step": 19413
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8201332092285156,
+      "learning_rate": 4.487073750991277e-05,
+      "loss": 1.3087,
+      "step": 19414
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.555067300796509,
+      "learning_rate": 4.487047316944224e-05,
+      "loss": 1.3896,
+      "step": 19415
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0856051445007324,
+      "learning_rate": 4.4870208828971716e-05,
+      "loss": 1.228,
+      "step": 19416
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.6119509935379028,
+      "learning_rate": 4.486994448850119e-05,
+      "loss": 1.1451,
+      "step": 19417
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.252721071243286,
+      "learning_rate": 4.486968014803067e-05,
+      "loss": 1.2349,
+      "step": 19418
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8976370096206665,
+      "learning_rate": 4.4869415807560137e-05,
+      "loss": 1.1957,
+      "step": 19419
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.5051424503326416,
+      "learning_rate": 4.486915146708962e-05,
+      "loss": 1.2916,
+      "step": 19420
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.4538605213165283,
+      "learning_rate": 4.4868887126619084e-05,
+      "loss": 2.098,
+      "step": 19421
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.985826015472412,
+      "learning_rate": 4.4868622786148564e-05,
+      "loss": 1.3357,
+      "step": 19422
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.98153018951416,
+      "learning_rate": 4.486835844567804e-05,
+      "loss": 1.3555,
+      "step": 19423
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.179598569869995,
+      "learning_rate": 4.486809410520751e-05,
+      "loss": 1.3249,
+      "step": 19424
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.078749656677246,
+      "learning_rate": 4.4867829764736985e-05,
+      "loss": 1.3465,
+      "step": 19425
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.8043248653411865,
+      "learning_rate": 4.486756542426646e-05,
+      "loss": 1.3128,
+      "step": 19426
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0143675804138184,
+      "learning_rate": 4.486730108379593e-05,
+      "loss": 1.1765,
+      "step": 19427
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.4886959791183472,
+      "learning_rate": 4.4867036743325406e-05,
+      "loss": 1.1129,
+      "step": 19428
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.1324706077575684,
+      "learning_rate": 4.486677240285488e-05,
+      "loss": 1.3538,
+      "step": 19429
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.902435541152954,
+      "learning_rate": 4.486650806238435e-05,
+      "loss": 1.3044,
+      "step": 19430
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7543118000030518,
+      "learning_rate": 4.4866243721913826e-05,
+      "loss": 1.2559,
+      "step": 19431
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.622443437576294,
+      "learning_rate": 4.48659793814433e-05,
+      "loss": 1.265,
+      "step": 19432
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8080627918243408,
+      "learning_rate": 4.486571504097278e-05,
+      "loss": 1.1395,
+      "step": 19433
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7389060258865356,
+      "learning_rate": 4.486545070050225e-05,
+      "loss": 1.2424,
+      "step": 19434
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.2800936698913574,
+      "learning_rate": 4.486518636003173e-05,
+      "loss": 1.3745,
+      "step": 19435
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.003380537033081,
+      "learning_rate": 4.4864922019561194e-05,
+      "loss": 1.3888,
+      "step": 19436
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.272517681121826,
+      "learning_rate": 4.4864657679090674e-05,
+      "loss": 1.3901,
+      "step": 19437
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.7932705879211426,
+      "learning_rate": 4.486439333862015e-05,
+      "loss": 1.403,
+      "step": 19438
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9293233156204224,
+      "learning_rate": 4.486412899814962e-05,
+      "loss": 1.238,
+      "step": 19439
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.065821647644043,
+      "learning_rate": 4.4863864657679095e-05,
+      "loss": 1.1295,
+      "step": 19440
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5805201530456543,
+      "learning_rate": 4.486360031720856e-05,
+      "loss": 1.1849,
+      "step": 19441
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.886168122291565,
+      "learning_rate": 4.486333597673804e-05,
+      "loss": 1.2893,
+      "step": 19442
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.400926113128662,
+      "learning_rate": 4.4863071636267516e-05,
+      "loss": 1.4633,
+      "step": 19443
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9270806312561035,
+      "learning_rate": 4.486280729579699e-05,
+      "loss": 2.3712,
+      "step": 19444
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.623645067214966,
+      "learning_rate": 4.486254295532646e-05,
+      "loss": 1.2738,
+      "step": 19445
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.248786687850952,
+      "learning_rate": 4.4862278614855937e-05,
+      "loss": 1.411,
+      "step": 19446
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.2885019779205322,
+      "learning_rate": 4.486201427438541e-05,
+      "loss": 1.4698,
+      "step": 19447
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9740331172943115,
+      "learning_rate": 4.486174993391489e-05,
+      "loss": 1.3196,
+      "step": 19448
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.979546070098877,
+      "learning_rate": 4.486148559344436e-05,
+      "loss": 1.4493,
+      "step": 19449
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1974215507507324,
+      "learning_rate": 4.486122125297383e-05,
+      "loss": 1.31,
+      "step": 19450
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7024933099746704,
+      "learning_rate": 4.4860956912503304e-05,
+      "loss": 1.3009,
+      "step": 19451
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8415395021438599,
+      "learning_rate": 4.486069257203278e-05,
+      "loss": 1.2265,
+      "step": 19452
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.4508743286132812,
+      "learning_rate": 4.486042823156226e-05,
+      "loss": 1.449,
+      "step": 19453
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.6167500019073486,
+      "learning_rate": 4.4860163891091725e-05,
+      "loss": 1.286,
+      "step": 19454
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.6951254606246948,
+      "learning_rate": 4.4859899550621205e-05,
+      "loss": 1.1732,
+      "step": 19455
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7635180950164795,
+      "learning_rate": 4.485963521015067e-05,
+      "loss": 1.2228,
+      "step": 19456
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.3727986812591553,
+      "learning_rate": 4.485937086968015e-05,
+      "loss": 1.4762,
+      "step": 19457
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.757432460784912,
+      "learning_rate": 4.4859106529209626e-05,
+      "loss": 1.4056,
+      "step": 19458
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.7506699562072754,
+      "learning_rate": 4.48588421887391e-05,
+      "loss": 1.2897,
+      "step": 19459
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.20986008644104,
+      "learning_rate": 4.485857784826857e-05,
+      "loss": 1.4378,
+      "step": 19460
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.175673007965088,
+      "learning_rate": 4.485831350779805e-05,
+      "loss": 1.3346,
+      "step": 19461
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1499533653259277,
+      "learning_rate": 4.485804916732752e-05,
+      "loss": 1.3472,
+      "step": 19462
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8519933223724365,
+      "learning_rate": 4.4857784826856994e-05,
+      "loss": 1.1981,
+      "step": 19463
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5772666931152344,
+      "learning_rate": 4.485752048638647e-05,
+      "loss": 1.1749,
+      "step": 19464
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.829308271408081,
+      "learning_rate": 4.485725614591594e-05,
+      "loss": 1.3666,
+      "step": 19465
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.760213613510132,
+      "learning_rate": 4.4856991805445415e-05,
+      "loss": 1.4393,
+      "step": 19466
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.733671188354492,
+      "learning_rate": 4.485672746497489e-05,
+      "loss": 1.4629,
+      "step": 19467
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8042243719100952,
+      "learning_rate": 4.485646312450437e-05,
+      "loss": 1.2239,
+      "step": 19468
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7505074739456177,
+      "learning_rate": 4.4856198784033835e-05,
+      "loss": 1.3186,
+      "step": 19469
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.883164405822754,
+      "learning_rate": 4.4855934443563316e-05,
+      "loss": 1.2461,
+      "step": 19470
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.994161367416382,
+      "learning_rate": 4.485567010309278e-05,
+      "loss": 1.6423,
+      "step": 19471
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.4511762857437134,
+      "learning_rate": 4.485540576262226e-05,
+      "loss": 1.6177,
+      "step": 19472
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.701138973236084,
+      "learning_rate": 4.4855141422151737e-05,
+      "loss": 1.4546,
+      "step": 19473
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.2249104976654053,
+      "learning_rate": 4.485487708168121e-05,
+      "loss": 1.5039,
+      "step": 19474
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.2188384532928467,
+      "learning_rate": 4.4854612741210684e-05,
+      "loss": 1.2984,
+      "step": 19475
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7743604183197021,
+      "learning_rate": 4.485434840074015e-05,
+      "loss": 1.3116,
+      "step": 19476
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5755031108856201,
+      "learning_rate": 4.485408406026963e-05,
+      "loss": 1.2017,
+      "step": 19477
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.081815719604492,
+      "learning_rate": 4.4853819719799104e-05,
+      "loss": 1.3503,
+      "step": 19478
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8347703218460083,
+      "learning_rate": 4.485355537932858e-05,
+      "loss": 1.2571,
+      "step": 19479
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.4626927375793457,
+      "learning_rate": 4.485329103885805e-05,
+      "loss": 1.278,
+      "step": 19480
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.6981512308120728,
+      "learning_rate": 4.4853026698387525e-05,
+      "loss": 1.2257,
+      "step": 19481
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8701767921447754,
+      "learning_rate": 4.4852762357917e-05,
+      "loss": 1.3425,
+      "step": 19482
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0629541873931885,
+      "learning_rate": 4.485249801744648e-05,
+      "loss": 1.4299,
+      "step": 19483
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7471022605895996,
+      "learning_rate": 4.4852233676975946e-05,
+      "loss": 1.4314,
+      "step": 19484
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.477726936340332,
+      "learning_rate": 4.485196933650542e-05,
+      "loss": 1.4087,
+      "step": 19485
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9239002466201782,
+      "learning_rate": 4.485170499603489e-05,
+      "loss": 1.2378,
+      "step": 19486
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8721624612808228,
+      "learning_rate": 4.4851440655564367e-05,
+      "loss": 1.3193,
+      "step": 19487
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9649465084075928,
+      "learning_rate": 4.485117631509385e-05,
+      "loss": 1.2468,
+      "step": 19488
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0481224060058594,
+      "learning_rate": 4.4850911974623314e-05,
+      "loss": 1.3602,
+      "step": 19489
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1750895977020264,
+      "learning_rate": 4.4850647634152794e-05,
+      "loss": 1.3931,
+      "step": 19490
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.387603998184204,
+      "learning_rate": 4.485038329368226e-05,
+      "loss": 1.3259,
+      "step": 19491
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.1732497215270996,
+      "learning_rate": 4.485011895321174e-05,
+      "loss": 1.344,
+      "step": 19492
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0899946689605713,
+      "learning_rate": 4.4849854612741215e-05,
+      "loss": 1.5062,
+      "step": 19493
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0579309463500977,
+      "learning_rate": 4.484959027227069e-05,
+      "loss": 1.3054,
+      "step": 19494
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.852632999420166,
+      "learning_rate": 4.484932593180016e-05,
+      "loss": 1.2119,
+      "step": 19495
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8343534469604492,
+      "learning_rate": 4.4849061591329635e-05,
+      "loss": 1.2458,
+      "step": 19496
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0990982055664062,
+      "learning_rate": 4.484879725085911e-05,
+      "loss": 1.1461,
+      "step": 19497
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1926333904266357,
+      "learning_rate": 4.484853291038858e-05,
+      "loss": 1.4246,
+      "step": 19498
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.051920175552368,
+      "learning_rate": 4.4848268569918056e-05,
+      "loss": 1.3347,
+      "step": 19499
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.199821710586548,
+      "learning_rate": 4.484800422944753e-05,
+      "loss": 1.4082,
+      "step": 19500
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.4168717861175537,
+      "learning_rate": 4.4847739888977e-05,
+      "loss": 1.4148,
+      "step": 19501
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.243117094039917,
+      "learning_rate": 4.484747554850648e-05,
+      "loss": 1.4063,
+      "step": 19502
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1117498874664307,
+      "learning_rate": 4.484721120803596e-05,
+      "loss": 1.2594,
+      "step": 19503
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.823007583618164,
+      "learning_rate": 4.4846946867565424e-05,
+      "loss": 1.2118,
+      "step": 19504
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1320183277130127,
+      "learning_rate": 4.4846682527094904e-05,
+      "loss": 1.3562,
+      "step": 19505
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.3421850204467773,
+      "learning_rate": 4.484641818662437e-05,
+      "loss": 1.2675,
+      "step": 19506
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.5259692668914795,
+      "learning_rate": 4.484615384615385e-05,
+      "loss": 1.0937,
+      "step": 19507
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.815427303314209,
+      "learning_rate": 4.4845889505683325e-05,
+      "loss": 1.2935,
+      "step": 19508
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9062341451644897,
+      "learning_rate": 4.48456251652128e-05,
+      "loss": 1.2985,
+      "step": 19509
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0381524562835693,
+      "learning_rate": 4.484536082474227e-05,
+      "loss": 1.4402,
+      "step": 19510
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.67019522190094,
+      "learning_rate": 4.4845096484271746e-05,
+      "loss": 1.3478,
+      "step": 19511
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.727761149406433,
+      "learning_rate": 4.484483214380122e-05,
+      "loss": 1.2593,
+      "step": 19512
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8624194860458374,
+      "learning_rate": 4.484456780333069e-05,
+      "loss": 1.2223,
+      "step": 19513
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.6084283590316772,
+      "learning_rate": 4.4844303462860166e-05,
+      "loss": 1.2296,
+      "step": 19514
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9499093294143677,
+      "learning_rate": 4.484403912238964e-05,
+      "loss": 1.3666,
+      "step": 19515
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.175250768661499,
+      "learning_rate": 4.4843774781919114e-05,
+      "loss": 1.2711,
+      "step": 19516
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9544496536254883,
+      "learning_rate": 4.484351044144859e-05,
+      "loss": 1.2586,
+      "step": 19517
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.6483389139175415,
+      "learning_rate": 4.484324610097806e-05,
+      "loss": 1.2148,
+      "step": 19518
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5501691102981567,
+      "learning_rate": 4.4842981760507534e-05,
+      "loss": 1.1533,
+      "step": 19519
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.0151615142822266,
+      "learning_rate": 4.4842717420037015e-05,
+      "loss": 1.339,
+      "step": 19520
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.652399778366089,
+      "learning_rate": 4.484245307956648e-05,
+      "loss": 1.2939,
+      "step": 19521
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.99347722530365,
+      "learning_rate": 4.4842188739095955e-05,
+      "loss": 1.3044,
+      "step": 19522
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.195204019546509,
+      "learning_rate": 4.484192439862543e-05,
+      "loss": 1.3339,
+      "step": 19523
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0153205394744873,
+      "learning_rate": 4.48416600581549e-05,
+      "loss": 1.3727,
+      "step": 19524
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1362228393554688,
+      "learning_rate": 4.484139571768438e-05,
+      "loss": 1.3342,
+      "step": 19525
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.879409670829773,
+      "learning_rate": 4.484113137721385e-05,
+      "loss": 1.2456,
+      "step": 19526
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8958232402801514,
+      "learning_rate": 4.484086703674333e-05,
+      "loss": 1.3628,
+      "step": 19527
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1731607913970947,
+      "learning_rate": 4.4840602696272796e-05,
+      "loss": 1.2548,
+      "step": 19528
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9690337181091309,
+      "learning_rate": 4.484033835580228e-05,
+      "loss": 1.3223,
+      "step": 19529
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.04302716255188,
+      "learning_rate": 4.484007401533175e-05,
+      "loss": 1.3967,
+      "step": 19530
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8524376153945923,
+      "learning_rate": 4.4839809674861224e-05,
+      "loss": 1.2165,
+      "step": 19531
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.189791440963745,
+      "learning_rate": 4.48395453343907e-05,
+      "loss": 1.2723,
+      "step": 19532
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9977463483810425,
+      "learning_rate": 4.483928099392017e-05,
+      "loss": 1.2847,
+      "step": 19533
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5809420347213745,
+      "learning_rate": 4.4839016653449645e-05,
+      "loss": 1.286,
+      "step": 19534
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7334957122802734,
+      "learning_rate": 4.483875231297912e-05,
+      "loss": 1.1876,
+      "step": 19535
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8814048767089844,
+      "learning_rate": 4.483848797250859e-05,
+      "loss": 1.394,
+      "step": 19536
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.337937116622925,
+      "learning_rate": 4.4838223632038065e-05,
+      "loss": 1.4496,
+      "step": 19537
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.708957552909851,
+      "learning_rate": 4.483795929156754e-05,
+      "loss": 1.3666,
+      "step": 19538
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8063745498657227,
+      "learning_rate": 4.483769495109701e-05,
+      "loss": 1.2386,
+      "step": 19539
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.3460566997528076,
+      "learning_rate": 4.483743061062649e-05,
+      "loss": 1.2403,
+      "step": 19540
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.268456220626831,
+      "learning_rate": 4.483716627015596e-05,
+      "loss": 1.3422,
+      "step": 19541
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7635308504104614,
+      "learning_rate": 4.483690192968544e-05,
+      "loss": 1.2934,
+      "step": 19542
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.804312825202942,
+      "learning_rate": 4.483663758921491e-05,
+      "loss": 1.358,
+      "step": 19543
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7719887495040894,
+      "learning_rate": 4.483637324874439e-05,
+      "loss": 1.2184,
+      "step": 19544
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.6865054368972778,
+      "learning_rate": 4.483610890827386e-05,
+      "loss": 1.2973,
+      "step": 19545
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9493569135665894,
+      "learning_rate": 4.4835844567803334e-05,
+      "loss": 1.315,
+      "step": 19546
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.2315711975097656,
+      "learning_rate": 4.483558022733281e-05,
+      "loss": 1.466,
+      "step": 19547
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0255777835845947,
+      "learning_rate": 4.4835315886862275e-05,
+      "loss": 1.3964,
+      "step": 19548
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7616959810256958,
+      "learning_rate": 4.4835051546391755e-05,
+      "loss": 1.2871,
+      "step": 19549
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.9059529304504395,
+      "learning_rate": 4.483478720592123e-05,
+      "loss": 1.4379,
+      "step": 19550
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.705665111541748,
+      "learning_rate": 4.48345228654507e-05,
+      "loss": 1.1158,
+      "step": 19551
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.9872629642486572,
+      "learning_rate": 4.4834258524980176e-05,
+      "loss": 1.3022,
+      "step": 19552
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0199923515319824,
+      "learning_rate": 4.483399418450965e-05,
+      "loss": 1.2932,
+      "step": 19553
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0424652099609375,
+      "learning_rate": 4.483372984403912e-05,
+      "loss": 1.2998,
+      "step": 19554
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8384990692138672,
+      "learning_rate": 4.48334655035686e-05,
+      "loss": 1.2789,
+      "step": 19555
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0284523963928223,
+      "learning_rate": 4.483320116309807e-05,
+      "loss": 1.4432,
+      "step": 19556
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9112292528152466,
+      "learning_rate": 4.483293682262755e-05,
+      "loss": 1.2403,
+      "step": 19557
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9865998029708862,
+      "learning_rate": 4.483267248215702e-05,
+      "loss": 1.305,
+      "step": 19558
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9684616327285767,
+      "learning_rate": 4.483240814168649e-05,
+      "loss": 1.2452,
+      "step": 19559
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.2624456882476807,
+      "learning_rate": 4.483214380121597e-05,
+      "loss": 1.3383,
+      "step": 19560
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.4591829776763916,
+      "learning_rate": 4.483187946074544e-05,
+      "loss": 1.4638,
+      "step": 19561
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.3243746757507324,
+      "learning_rate": 4.483161512027492e-05,
+      "loss": 1.4087,
+      "step": 19562
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8095899820327759,
+      "learning_rate": 4.4831350779804385e-05,
+      "loss": 1.3104,
+      "step": 19563
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.046278238296509,
+      "learning_rate": 4.4831086439333865e-05,
+      "loss": 1.3348,
+      "step": 19564
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7795195579528809,
+      "learning_rate": 4.483082209886334e-05,
+      "loss": 1.4968,
+      "step": 19565
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.995771050453186,
+      "learning_rate": 4.483055775839281e-05,
+      "loss": 1.352,
+      "step": 19566
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1145389080047607,
+      "learning_rate": 4.4830293417922286e-05,
+      "loss": 1.3295,
+      "step": 19567
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.187133312225342,
+      "learning_rate": 4.483002907745176e-05,
+      "loss": 1.3779,
+      "step": 19568
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7673810720443726,
+      "learning_rate": 4.482976473698123e-05,
+      "loss": 1.1299,
+      "step": 19569
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7729252576828003,
+      "learning_rate": 4.482950039651071e-05,
+      "loss": 1.1238,
+      "step": 19570
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.1095902919769287,
+      "learning_rate": 4.482923605604018e-05,
+      "loss": 1.494,
+      "step": 19571
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.6589736938476562,
+      "learning_rate": 4.4828971715569654e-05,
+      "loss": 1.666,
+      "step": 19572
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0586860179901123,
+      "learning_rate": 4.482870737509913e-05,
+      "loss": 1.3432,
+      "step": 19573
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5149792432785034,
+      "learning_rate": 4.48284430346286e-05,
+      "loss": 1.2274,
+      "step": 19574
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.78371000289917,
+      "learning_rate": 4.482817869415808e-05,
+      "loss": 1.2837,
+      "step": 19575
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.4179739952087402,
+      "learning_rate": 4.482791435368755e-05,
+      "loss": 1.3819,
+      "step": 19576
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1149210929870605,
+      "learning_rate": 4.482765001321703e-05,
+      "loss": 1.3497,
+      "step": 19577
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.5326521396636963,
+      "learning_rate": 4.4827385672746495e-05,
+      "loss": 1.464,
+      "step": 19578
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.95244562625885,
+      "learning_rate": 4.4827121332275976e-05,
+      "loss": 1.2843,
+      "step": 19579
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.3902835845947266,
+      "learning_rate": 4.482685699180545e-05,
+      "loss": 1.3341,
+      "step": 19580
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0268633365631104,
+      "learning_rate": 4.482659265133492e-05,
+      "loss": 1.4032,
+      "step": 19581
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.793129563331604,
+      "learning_rate": 4.4826328310864396e-05,
+      "loss": 1.2581,
+      "step": 19582
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0416197776794434,
+      "learning_rate": 4.482606397039387e-05,
+      "loss": 1.3493,
+      "step": 19583
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6348210573196411,
+      "learning_rate": 4.4825799629923344e-05,
+      "loss": 1.1471,
+      "step": 19584
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.795670986175537,
+      "learning_rate": 4.482553528945282e-05,
+      "loss": 1.1449,
+      "step": 19585
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7945168018341064,
+      "learning_rate": 4.482527094898229e-05,
+      "loss": 1.2235,
+      "step": 19586
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9967659711837769,
+      "learning_rate": 4.4825006608511764e-05,
+      "loss": 1.3553,
+      "step": 19587
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.785787582397461,
+      "learning_rate": 4.482474226804124e-05,
+      "loss": 1.2783,
+      "step": 19588
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.283588409423828,
+      "learning_rate": 4.482447792757071e-05,
+      "loss": 1.4488,
+      "step": 19589
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8390940427780151,
+      "learning_rate": 4.482421358710019e-05,
+      "loss": 1.286,
+      "step": 19590
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.218482732772827,
+      "learning_rate": 4.482394924662966e-05,
+      "loss": 1.3063,
+      "step": 19591
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.4676096439361572,
+      "learning_rate": 4.482368490615914e-05,
+      "loss": 1.4445,
+      "step": 19592
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.722224473953247,
+      "learning_rate": 4.4823420565688606e-05,
+      "loss": 1.3809,
+      "step": 19593
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6730048656463623,
+      "learning_rate": 4.482315622521808e-05,
+      "loss": 1.1534,
+      "step": 19594
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8761783838272095,
+      "learning_rate": 4.482289188474756e-05,
+      "loss": 1.4548,
+      "step": 19595
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.022587299346924,
+      "learning_rate": 4.4822627544277026e-05,
+      "loss": 1.3086,
+      "step": 19596
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8217889070510864,
+      "learning_rate": 4.482236320380651e-05,
+      "loss": 1.2831,
+      "step": 19597
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9359257221221924,
+      "learning_rate": 4.4822098863335974e-05,
+      "loss": 1.397,
+      "step": 19598
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8676819801330566,
+      "learning_rate": 4.4821834522865454e-05,
+      "loss": 1.2258,
+      "step": 19599
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.7394471168518066,
+      "learning_rate": 4.482157018239493e-05,
+      "loss": 1.4148,
+      "step": 19600
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.4670512676239014,
+      "learning_rate": 4.48213058419244e-05,
+      "loss": 1.4457,
+      "step": 19601
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5779696702957153,
+      "learning_rate": 4.4821041501453875e-05,
+      "loss": 1.209,
+      "step": 19602
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.056147575378418,
+      "learning_rate": 4.482077716098335e-05,
+      "loss": 1.4303,
+      "step": 19603
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.994632601737976,
+      "learning_rate": 4.482051282051282e-05,
+      "loss": 1.3167,
+      "step": 19604
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.634329319000244,
+      "learning_rate": 4.4820248480042295e-05,
+      "loss": 1.2728,
+      "step": 19605
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8657691478729248,
+      "learning_rate": 4.481998413957177e-05,
+      "loss": 1.2871,
+      "step": 19606
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7917780876159668,
+      "learning_rate": 4.481971979910124e-05,
+      "loss": 1.2099,
+      "step": 19607
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.120419502258301,
+      "learning_rate": 4.4819455458630716e-05,
+      "loss": 1.3102,
+      "step": 19608
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0708353519439697,
+      "learning_rate": 4.481919111816019e-05,
+      "loss": 1.2847,
+      "step": 19609
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8047508001327515,
+      "learning_rate": 4.481892677768967e-05,
+      "loss": 1.2162,
+      "step": 19610
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0790090560913086,
+      "learning_rate": 4.481866243721914e-05,
+      "loss": 1.3208,
+      "step": 19611
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.3672523498535156,
+      "learning_rate": 4.481839809674862e-05,
+      "loss": 1.3211,
+      "step": 19612
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8099863529205322,
+      "learning_rate": 4.4818133756278084e-05,
+      "loss": 1.2865,
+      "step": 19613
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0155444145202637,
+      "learning_rate": 4.4817869415807564e-05,
+      "loss": 1.3041,
+      "step": 19614
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.061368942260742,
+      "learning_rate": 4.481760507533704e-05,
+      "loss": 1.2589,
+      "step": 19615
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5447022914886475,
+      "learning_rate": 4.481734073486651e-05,
+      "loss": 1.0351,
+      "step": 19616
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.99787175655365,
+      "learning_rate": 4.4817076394395985e-05,
+      "loss": 1.2216,
+      "step": 19617
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6418204307556152,
+      "learning_rate": 4.481681205392546e-05,
+      "loss": 1.0801,
+      "step": 19618
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0280532836914062,
+      "learning_rate": 4.481654771345493e-05,
+      "loss": 1.2447,
+      "step": 19619
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8366018533706665,
+      "learning_rate": 4.4816283372984406e-05,
+      "loss": 1.2884,
+      "step": 19620
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9576222896575928,
+      "learning_rate": 4.481601903251388e-05,
+      "loss": 1.2633,
+      "step": 19621
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.503296136856079,
+      "learning_rate": 4.481575469204335e-05,
+      "loss": 1.1711,
+      "step": 19622
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7536500692367554,
+      "learning_rate": 4.4815490351572826e-05,
+      "loss": 1.3272,
+      "step": 19623
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1360597610473633,
+      "learning_rate": 4.48152260111023e-05,
+      "loss": 1.4427,
+      "step": 19624
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.97330904006958,
+      "learning_rate": 4.481496167063178e-05,
+      "loss": 1.245,
+      "step": 19625
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.625332832336426,
+      "learning_rate": 4.481469733016125e-05,
+      "loss": 1.3933,
+      "step": 19626
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.003950357437134,
+      "learning_rate": 4.481443298969073e-05,
+      "loss": 1.2461,
+      "step": 19627
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8985320329666138,
+      "learning_rate": 4.4814168649220194e-05,
+      "loss": 1.2796,
+      "step": 19628
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5948717594146729,
+      "learning_rate": 4.4813904308749675e-05,
+      "loss": 1.2395,
+      "step": 19629
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8889631032943726,
+      "learning_rate": 4.481363996827915e-05,
+      "loss": 1.3208,
+      "step": 19630
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5543872117996216,
+      "learning_rate": 4.4813375627808615e-05,
+      "loss": 1.1813,
+      "step": 19631
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.023261785507202,
+      "learning_rate": 4.4813111287338095e-05,
+      "loss": 1.4092,
+      "step": 19632
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2453296184539795,
+      "learning_rate": 4.481284694686756e-05,
+      "loss": 1.4529,
+      "step": 19633
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6036598682403564,
+      "learning_rate": 4.481258260639704e-05,
+      "loss": 1.1647,
+      "step": 19634
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8083544969558716,
+      "learning_rate": 4.4812318265926516e-05,
+      "loss": 1.3435,
+      "step": 19635
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1401638984680176,
+      "learning_rate": 4.481205392545599e-05,
+      "loss": 1.3478,
+      "step": 19636
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.832443356513977,
+      "learning_rate": 4.481178958498546e-05,
+      "loss": 1.1837,
+      "step": 19637
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8234814405441284,
+      "learning_rate": 4.481152524451494e-05,
+      "loss": 2.1839,
+      "step": 19638
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.4397754669189453,
+      "learning_rate": 4.481126090404441e-05,
+      "loss": 1.3921,
+      "step": 19639
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5809085369110107,
+      "learning_rate": 4.4810996563573884e-05,
+      "loss": 1.2105,
+      "step": 19640
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6767903566360474,
+      "learning_rate": 4.481073222310336e-05,
+      "loss": 1.1671,
+      "step": 19641
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1763973236083984,
+      "learning_rate": 4.481046788263283e-05,
+      "loss": 1.2704,
+      "step": 19642
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1712656021118164,
+      "learning_rate": 4.4810203542162305e-05,
+      "loss": 1.3112,
+      "step": 19643
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8175227642059326,
+      "learning_rate": 4.480993920169178e-05,
+      "loss": 1.2552,
+      "step": 19644
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.078167200088501,
+      "learning_rate": 4.480967486122126e-05,
+      "loss": 1.2474,
+      "step": 19645
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0015363693237305,
+      "learning_rate": 4.4809410520750725e-05,
+      "loss": 1.3121,
+      "step": 19646
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.6996893882751465,
+      "learning_rate": 4.4809146180280206e-05,
+      "loss": 1.4509,
+      "step": 19647
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4878829717636108,
+      "learning_rate": 4.480888183980967e-05,
+      "loss": 1.1475,
+      "step": 19648
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.74752140045166,
+      "learning_rate": 4.480861749933915e-05,
+      "loss": 1.2394,
+      "step": 19649
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.6137499809265137,
+      "learning_rate": 4.4808353158868626e-05,
+      "loss": 1.4032,
+      "step": 19650
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6387869119644165,
+      "learning_rate": 4.48080888183981e-05,
+      "loss": 1.3371,
+      "step": 19651
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.212766647338867,
+      "learning_rate": 4.4807824477927573e-05,
+      "loss": 1.3675,
+      "step": 19652
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9057319164276123,
+      "learning_rate": 4.480756013745705e-05,
+      "loss": 1.3212,
+      "step": 19653
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.422418475151062,
+      "learning_rate": 4.480729579698652e-05,
+      "loss": 1.0831,
+      "step": 19654
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7779031991958618,
+      "learning_rate": 4.4807031456515994e-05,
+      "loss": 1.2495,
+      "step": 19655
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.835635781288147,
+      "learning_rate": 4.480676711604547e-05,
+      "loss": 1.1892,
+      "step": 19656
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.946498990058899,
+      "learning_rate": 4.480650277557494e-05,
+      "loss": 1.3425,
+      "step": 19657
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4194926023483276,
+      "learning_rate": 4.4806238435104415e-05,
+      "loss": 1.1834,
+      "step": 19658
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8889490365982056,
+      "learning_rate": 4.480597409463389e-05,
+      "loss": 1.3641,
+      "step": 19659
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.4394946098327637,
+      "learning_rate": 4.480570975416337e-05,
+      "loss": 1.4176,
+      "step": 19660
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.403265953063965,
+      "learning_rate": 4.4805445413692836e-05,
+      "loss": 1.4644,
+      "step": 19661
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.4576849937438965,
+      "learning_rate": 4.4805181073222316e-05,
+      "loss": 1.2885,
+      "step": 19662
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.128666639328003,
+      "learning_rate": 4.480491673275178e-05,
+      "loss": 1.326,
+      "step": 19663
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.788472890853882,
+      "learning_rate": 4.480465239228126e-05,
+      "loss": 1.3917,
+      "step": 19664
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.338827133178711,
+      "learning_rate": 4.480438805181074e-05,
+      "loss": 1.4648,
+      "step": 19665
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.4104819297790527,
+      "learning_rate": 4.4804123711340203e-05,
+      "loss": 1.4221,
+      "step": 19666
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.028618097305298,
+      "learning_rate": 4.4803859370869684e-05,
+      "loss": 1.3357,
+      "step": 19667
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.974151611328125,
+      "learning_rate": 4.480359503039915e-05,
+      "loss": 1.31,
+      "step": 19668
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1103386878967285,
+      "learning_rate": 4.480333068992863e-05,
+      "loss": 1.3218,
+      "step": 19669
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.5550432205200195,
+      "learning_rate": 4.4803066349458105e-05,
+      "loss": 1.2572,
+      "step": 19670
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5881006717681885,
+      "learning_rate": 4.480280200898758e-05,
+      "loss": 1.2246,
+      "step": 19671
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7831108570098877,
+      "learning_rate": 4.480253766851705e-05,
+      "loss": 1.2734,
+      "step": 19672
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.370614528656006,
+      "learning_rate": 4.4802273328046525e-05,
+      "loss": 1.216,
+      "step": 19673
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.233865737915039,
+      "learning_rate": 4.4802008987576e-05,
+      "loss": 1.2703,
+      "step": 19674
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.154013156890869,
+      "learning_rate": 4.480174464710548e-05,
+      "loss": 1.4104,
+      "step": 19675
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6404343843460083,
+      "learning_rate": 4.4801480306634946e-05,
+      "loss": 1.3203,
+      "step": 19676
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7899051904678345,
+      "learning_rate": 4.480121596616442e-05,
+      "loss": 1.1919,
+      "step": 19677
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0714316368103027,
+      "learning_rate": 4.480095162569389e-05,
+      "loss": 1.2169,
+      "step": 19678
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5834848880767822,
+      "learning_rate": 4.480068728522337e-05,
+      "loss": 1.1869,
+      "step": 19679
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.078671455383301,
+      "learning_rate": 4.480042294475285e-05,
+      "loss": 1.4521,
+      "step": 19680
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8836688995361328,
+      "learning_rate": 4.4800158604282314e-05,
+      "loss": 1.2916,
+      "step": 19681
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1447601318359375,
+      "learning_rate": 4.4799894263811794e-05,
+      "loss": 1.3585,
+      "step": 19682
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.992504596710205,
+      "learning_rate": 4.479962992334126e-05,
+      "loss": 1.4228,
+      "step": 19683
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9974435567855835,
+      "learning_rate": 4.479936558287074e-05,
+      "loss": 1.1472,
+      "step": 19684
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7217354774475098,
+      "learning_rate": 4.4799101242400215e-05,
+      "loss": 1.1398,
+      "step": 19685
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6640639305114746,
+      "learning_rate": 4.479883690192969e-05,
+      "loss": 1.273,
+      "step": 19686
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.4587936401367188,
+      "learning_rate": 4.479857256145916e-05,
+      "loss": 1.4995,
+      "step": 19687
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.848068356513977,
+      "learning_rate": 4.4798308220988636e-05,
+      "loss": 1.338,
+      "step": 19688
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.821587085723877,
+      "learning_rate": 4.479804388051811e-05,
+      "loss": 1.2351,
+      "step": 19689
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.238354206085205,
+      "learning_rate": 4.479777954004758e-05,
+      "loss": 1.5101,
+      "step": 19690
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9472719430923462,
+      "learning_rate": 4.4797515199577056e-05,
+      "loss": 1.2569,
+      "step": 19691
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0808494091033936,
+      "learning_rate": 4.479725085910653e-05,
+      "loss": 1.4501,
+      "step": 19692
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.001495361328125,
+      "learning_rate": 4.4796986518636003e-05,
+      "loss": 1.3639,
+      "step": 19693
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2133774757385254,
+      "learning_rate": 4.479672217816548e-05,
+      "loss": 1.2857,
+      "step": 19694
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7854565382003784,
+      "learning_rate": 4.479645783769496e-05,
+      "loss": 1.2578,
+      "step": 19695
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9698591232299805,
+      "learning_rate": 4.4796193497224424e-05,
+      "loss": 1.268,
+      "step": 19696
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.825249195098877,
+      "learning_rate": 4.4795929156753905e-05,
+      "loss": 1.1826,
+      "step": 19697
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.588602900505066,
+      "learning_rate": 4.479566481628337e-05,
+      "loss": 1.321,
+      "step": 19698
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1623759269714355,
+      "learning_rate": 4.479540047581285e-05,
+      "loss": 1.3438,
+      "step": 19699
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9911571741104126,
+      "learning_rate": 4.4795136135342325e-05,
+      "loss": 1.317,
+      "step": 19700
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.835992693901062,
+      "learning_rate": 4.47948717948718e-05,
+      "loss": 1.4432,
+      "step": 19701
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.109523057937622,
+      "learning_rate": 4.479460745440127e-05,
+      "loss": 1.1629,
+      "step": 19702
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.997902750968933,
+      "learning_rate": 4.479434311393074e-05,
+      "loss": 1.3846,
+      "step": 19703
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0116117000579834,
+      "learning_rate": 4.479407877346022e-05,
+      "loss": 1.2904,
+      "step": 19704
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9430991411209106,
+      "learning_rate": 4.479381443298969e-05,
+      "loss": 1.3719,
+      "step": 19705
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0975112915039062,
+      "learning_rate": 4.479355009251917e-05,
+      "loss": 1.4044,
+      "step": 19706
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.214747428894043,
+      "learning_rate": 4.479328575204864e-05,
+      "loss": 1.2833,
+      "step": 19707
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.3799548149108887,
+      "learning_rate": 4.4793021411578114e-05,
+      "loss": 1.5071,
+      "step": 19708
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.248619318008423,
+      "learning_rate": 4.479275707110759e-05,
+      "loss": 1.3503,
+      "step": 19709
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.965542197227478,
+      "learning_rate": 4.479249273063707e-05,
+      "loss": 1.2522,
+      "step": 19710
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8632915019989014,
+      "learning_rate": 4.4792228390166535e-05,
+      "loss": 1.1799,
+      "step": 19711
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6109379529953003,
+      "learning_rate": 4.479196404969601e-05,
+      "loss": 1.1792,
+      "step": 19712
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1255407333374023,
+      "learning_rate": 4.479169970922548e-05,
+      "loss": 1.3145,
+      "step": 19713
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0501883029937744,
+      "learning_rate": 4.4791435368754955e-05,
+      "loss": 1.3296,
+      "step": 19714
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.33401346206665,
+      "learning_rate": 4.4791171028284436e-05,
+      "loss": 1.4666,
+      "step": 19715
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9047099351882935,
+      "learning_rate": 4.47909066878139e-05,
+      "loss": 1.3469,
+      "step": 19716
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6915744543075562,
+      "learning_rate": 4.479064234734338e-05,
+      "loss": 1.2122,
+      "step": 19717
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8687314987182617,
+      "learning_rate": 4.479037800687285e-05,
+      "loss": 1.359,
+      "step": 19718
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9068862199783325,
+      "learning_rate": 4.479011366640233e-05,
+      "loss": 1.2365,
+      "step": 19719
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7869246006011963,
+      "learning_rate": 4.4789849325931803e-05,
+      "loss": 1.169,
+      "step": 19720
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6334365606307983,
+      "learning_rate": 4.478958498546128e-05,
+      "loss": 1.1514,
+      "step": 19721
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1661674976348877,
+      "learning_rate": 4.478932064499075e-05,
+      "loss": 1.29,
+      "step": 19722
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9005060195922852,
+      "learning_rate": 4.4789056304520224e-05,
+      "loss": 1.1699,
+      "step": 19723
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8302136659622192,
+      "learning_rate": 4.47887919640497e-05,
+      "loss": 1.1124,
+      "step": 19724
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.170053482055664,
+      "learning_rate": 4.478852762357917e-05,
+      "loss": 1.2862,
+      "step": 19725
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.899840235710144,
+      "learning_rate": 4.4788263283108645e-05,
+      "loss": 1.304,
+      "step": 19726
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5831376314163208,
+      "learning_rate": 4.478799894263812e-05,
+      "loss": 1.1695,
+      "step": 19727
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8669673204421997,
+      "learning_rate": 4.478773460216759e-05,
+      "loss": 1.1548,
+      "step": 19728
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0887911319732666,
+      "learning_rate": 4.4787470261697066e-05,
+      "loss": 1.2796,
+      "step": 19729
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0411641597747803,
+      "learning_rate": 4.4787205921226546e-05,
+      "loss": 1.297,
+      "step": 19730
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7286055088043213,
+      "learning_rate": 4.478694158075601e-05,
+      "loss": 1.3767,
+      "step": 19731
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.681370973587036,
+      "learning_rate": 4.478667724028549e-05,
+      "loss": 1.4328,
+      "step": 19732
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0413801670074463,
+      "learning_rate": 4.478641289981496e-05,
+      "loss": 1.2839,
+      "step": 19733
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.274768352508545,
+      "learning_rate": 4.478614855934444e-05,
+      "loss": 1.3155,
+      "step": 19734
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4005333185195923,
+      "learning_rate": 4.4785884218873914e-05,
+      "loss": 1.0254,
+      "step": 19735
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7894083261489868,
+      "learning_rate": 4.478561987840339e-05,
+      "loss": 1.2487,
+      "step": 19736
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.466566324234009,
+      "learning_rate": 4.478535553793286e-05,
+      "loss": 1.318,
+      "step": 19737
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.6641316413879395,
+      "learning_rate": 4.478509119746233e-05,
+      "loss": 2.3645,
+      "step": 19738
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1586709022521973,
+      "learning_rate": 4.478482685699181e-05,
+      "loss": 1.3533,
+      "step": 19739
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.124788999557495,
+      "learning_rate": 4.478456251652128e-05,
+      "loss": 1.5054,
+      "step": 19740
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9513723850250244,
+      "learning_rate": 4.4784298176050755e-05,
+      "loss": 1.3764,
+      "step": 19741
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.02013897895813,
+      "learning_rate": 4.478403383558023e-05,
+      "loss": 1.4004,
+      "step": 19742
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0124261379241943,
+      "learning_rate": 4.47837694951097e-05,
+      "loss": 1.3068,
+      "step": 19743
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.3688230514526367,
+      "learning_rate": 4.4783505154639176e-05,
+      "loss": 2.4265,
+      "step": 19744
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1251320838928223,
+      "learning_rate": 4.4783240814168656e-05,
+      "loss": 1.2936,
+      "step": 19745
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6949712038040161,
+      "learning_rate": 4.478297647369812e-05,
+      "loss": 1.1184,
+      "step": 19746
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.101621150970459,
+      "learning_rate": 4.4782712133227603e-05,
+      "loss": 1.2858,
+      "step": 19747
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5209771394729614,
+      "learning_rate": 4.478244779275707e-05,
+      "loss": 1.2478,
+      "step": 19748
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.683105707168579,
+      "learning_rate": 4.4782183452286544e-05,
+      "loss": 1.1958,
+      "step": 19749
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9519401788711548,
+      "learning_rate": 4.4781919111816024e-05,
+      "loss": 1.2922,
+      "step": 19750
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.952144742012024,
+      "learning_rate": 4.478165477134549e-05,
+      "loss": 1.1725,
+      "step": 19751
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.568863868713379,
+      "learning_rate": 4.478139043087497e-05,
+      "loss": 1.7295,
+      "step": 19752
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0621707439422607,
+      "learning_rate": 4.478112609040444e-05,
+      "loss": 1.2744,
+      "step": 19753
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7558887004852295,
+      "learning_rate": 4.478086174993392e-05,
+      "loss": 1.2874,
+      "step": 19754
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.908966302871704,
+      "learning_rate": 4.478059740946339e-05,
+      "loss": 1.2181,
+      "step": 19755
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.408696174621582,
+      "learning_rate": 4.4780333068992866e-05,
+      "loss": 1.506,
+      "step": 19756
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7865309715270996,
+      "learning_rate": 4.478006872852234e-05,
+      "loss": 1.2235,
+      "step": 19757
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6707769632339478,
+      "learning_rate": 4.477980438805181e-05,
+      "loss": 1.3201,
+      "step": 19758
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.751603603363037,
+      "learning_rate": 4.4779540047581286e-05,
+      "loss": 1.3053,
+      "step": 19759
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9741188287734985,
+      "learning_rate": 4.477927570711076e-05,
+      "loss": 1.3436,
+      "step": 19760
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7045960426330566,
+      "learning_rate": 4.477901136664023e-05,
+      "loss": 1.2114,
+      "step": 19761
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8483542203903198,
+      "learning_rate": 4.477874702616971e-05,
+      "loss": 1.1878,
+      "step": 19762
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6584563255310059,
+      "learning_rate": 4.477848268569918e-05,
+      "loss": 1.1425,
+      "step": 19763
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0585949420928955,
+      "learning_rate": 4.4778218345228654e-05,
+      "loss": 1.3454,
+      "step": 19764
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9828555583953857,
+      "learning_rate": 4.4777954004758134e-05,
+      "loss": 1.1717,
+      "step": 19765
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9585542678833008,
+      "learning_rate": 4.47776896642876e-05,
+      "loss": 1.3397,
+      "step": 19766
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.034646511077881,
+      "learning_rate": 4.477742532381708e-05,
+      "loss": 1.3622,
+      "step": 19767
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7979793548583984,
+      "learning_rate": 4.477716098334655e-05,
+      "loss": 1.2135,
+      "step": 19768
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7232171297073364,
+      "learning_rate": 4.477689664287603e-05,
+      "loss": 1.2724,
+      "step": 19769
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.101254463195801,
+      "learning_rate": 4.47766323024055e-05,
+      "loss": 1.3205,
+      "step": 19770
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0199992656707764,
+      "learning_rate": 4.4776367961934976e-05,
+      "loss": 1.2795,
+      "step": 19771
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7626385688781738,
+      "learning_rate": 4.477610362146445e-05,
+      "loss": 1.3428,
+      "step": 19772
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.5680993795394897,
+      "learning_rate": 4.477583928099392e-05,
+      "loss": 1.1308,
+      "step": 19773
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.134557008743286,
+      "learning_rate": 4.4775574940523397e-05,
+      "loss": 1.3337,
+      "step": 19774
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.234449625015259,
+      "learning_rate": 4.477531060005287e-05,
+      "loss": 1.3048,
+      "step": 19775
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9007179737091064,
+      "learning_rate": 4.4775046259582344e-05,
+      "loss": 1.2994,
+      "step": 19776
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4878854751586914,
+      "learning_rate": 4.477478191911182e-05,
+      "loss": 1.3968,
+      "step": 19777
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.914886474609375,
+      "learning_rate": 4.477451757864129e-05,
+      "loss": 1.2629,
+      "step": 19778
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4447858333587646,
+      "learning_rate": 4.4774253238170764e-05,
+      "loss": 1.3383,
+      "step": 19779
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1779065132141113,
+      "learning_rate": 4.4773988897700245e-05,
+      "loss": 1.4017,
+      "step": 19780
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.265446424484253,
+      "learning_rate": 4.477372455722971e-05,
+      "loss": 1.5033,
+      "step": 19781
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.668910026550293,
+      "learning_rate": 4.477346021675919e-05,
+      "loss": 1.337,
+      "step": 19782
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.210845947265625,
+      "learning_rate": 4.477319587628866e-05,
+      "loss": 1.8237,
+      "step": 19783
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.858041763305664,
+      "learning_rate": 4.477293153581813e-05,
+      "loss": 1.2502,
+      "step": 19784
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.232193946838379,
+      "learning_rate": 4.477266719534761e-05,
+      "loss": 1.3566,
+      "step": 19785
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1962554454803467,
+      "learning_rate": 4.477240285487708e-05,
+      "loss": 1.2999,
+      "step": 19786
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9965368509292603,
+      "learning_rate": 4.477213851440656e-05,
+      "loss": 1.2066,
+      "step": 19787
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6431102752685547,
+      "learning_rate": 4.4771874173936027e-05,
+      "loss": 1.3088,
+      "step": 19788
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.9703879356384277,
+      "learning_rate": 4.477160983346551e-05,
+      "loss": 1.3095,
+      "step": 19789
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4491493701934814,
+      "learning_rate": 4.477134549299498e-05,
+      "loss": 1.3148,
+      "step": 19790
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.144782781600952,
+      "learning_rate": 4.4771081152524454e-05,
+      "loss": 1.4016,
+      "step": 19791
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.422924518585205,
+      "learning_rate": 4.477081681205393e-05,
+      "loss": 1.5535,
+      "step": 19792
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.007931709289551,
+      "learning_rate": 4.47705524715834e-05,
+      "loss": 1.3166,
+      "step": 19793
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8180792331695557,
+      "learning_rate": 4.4770288131112875e-05,
+      "loss": 1.3087,
+      "step": 19794
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8477648496627808,
+      "learning_rate": 4.477002379064235e-05,
+      "loss": 1.1966,
+      "step": 19795
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6528682708740234,
+      "learning_rate": 4.476975945017182e-05,
+      "loss": 1.1894,
+      "step": 19796
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.19425368309021,
+      "learning_rate": 4.4769495109701295e-05,
+      "loss": 1.374,
+      "step": 19797
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9216169118881226,
+      "learning_rate": 4.476923076923077e-05,
+      "loss": 1.2943,
+      "step": 19798
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7183852195739746,
+      "learning_rate": 4.476896642876024e-05,
+      "loss": 1.4486,
+      "step": 19799
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.5440596342086792,
+      "learning_rate": 4.476870208828972e-05,
+      "loss": 1.1462,
+      "step": 19800
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6807578802108765,
+      "learning_rate": 4.476843774781919e-05,
+      "loss": 1.2004,
+      "step": 19801
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1303365230560303,
+      "learning_rate": 4.476817340734867e-05,
+      "loss": 1.419,
+      "step": 19802
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9450470209121704,
+      "learning_rate": 4.476790906687814e-05,
+      "loss": 1.4526,
+      "step": 19803
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.735687494277954,
+      "learning_rate": 4.476764472640762e-05,
+      "loss": 1.15,
+      "step": 19804
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4701526165008545,
+      "learning_rate": 4.476738038593709e-05,
+      "loss": 1.1725,
+      "step": 19805
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1192190647125244,
+      "learning_rate": 4.4767116045466564e-05,
+      "loss": 1.3538,
+      "step": 19806
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2072255611419678,
+      "learning_rate": 4.476685170499604e-05,
+      "loss": 1.303,
+      "step": 19807
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9251030683517456,
+      "learning_rate": 4.476658736452551e-05,
+      "loss": 1.4163,
+      "step": 19808
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7421015501022339,
+      "learning_rate": 4.4766323024054985e-05,
+      "loss": 1.1838,
+      "step": 19809
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.127608299255371,
+      "learning_rate": 4.476605868358446e-05,
+      "loss": 1.2522,
+      "step": 19810
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.112473487854004,
+      "learning_rate": 4.476579434311393e-05,
+      "loss": 1.3682,
+      "step": 19811
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.206110954284668,
+      "learning_rate": 4.4765530002643406e-05,
+      "loss": 1.3606,
+      "step": 19812
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.3683440685272217,
+      "learning_rate": 4.476526566217288e-05,
+      "loss": 1.3137,
+      "step": 19813
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.758830189704895,
+      "learning_rate": 4.476500132170235e-05,
+      "loss": 1.3238,
+      "step": 19814
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1899006366729736,
+      "learning_rate": 4.476473698123183e-05,
+      "loss": 1.1708,
+      "step": 19815
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.14398193359375,
+      "learning_rate": 4.47644726407613e-05,
+      "loss": 1.2143,
+      "step": 19816
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6632590293884277,
+      "learning_rate": 4.476420830029078e-05,
+      "loss": 1.2907,
+      "step": 19817
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.909954309463501,
+      "learning_rate": 4.476394395982025e-05,
+      "loss": 1.3476,
+      "step": 19818
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.090510368347168,
+      "learning_rate": 4.476367961934973e-05,
+      "loss": 1.4277,
+      "step": 19819
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.142996072769165,
+      "learning_rate": 4.47634152788792e-05,
+      "loss": 1.2814,
+      "step": 19820
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.768682599067688,
+      "learning_rate": 4.476315093840867e-05,
+      "loss": 1.2347,
+      "step": 19821
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2160017490386963,
+      "learning_rate": 4.476288659793815e-05,
+      "loss": 1.2586,
+      "step": 19822
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8946059942245483,
+      "learning_rate": 4.4762622257467615e-05,
+      "loss": 1.2798,
+      "step": 19823
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1211163997650146,
+      "learning_rate": 4.4762357916997095e-05,
+      "loss": 1.2787,
+      "step": 19824
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.231332778930664,
+      "learning_rate": 4.476209357652657e-05,
+      "loss": 1.4553,
+      "step": 19825
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.5587832927703857,
+      "learning_rate": 4.476182923605604e-05,
+      "loss": 1.3385,
+      "step": 19826
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0446112155914307,
+      "learning_rate": 4.4761564895585516e-05,
+      "loss": 1.168,
+      "step": 19827
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8423336744308472,
+      "learning_rate": 4.476130055511499e-05,
+      "loss": 1.3925,
+      "step": 19828
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7976353168487549,
+      "learning_rate": 4.476103621464446e-05,
+      "loss": 1.2597,
+      "step": 19829
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.232539653778076,
+      "learning_rate": 4.476077187417394e-05,
+      "loss": 1.0955,
+      "step": 19830
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.987046003341675,
+      "learning_rate": 4.476050753370341e-05,
+      "loss": 1.3086,
+      "step": 19831
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1913719177246094,
+      "learning_rate": 4.4760243193232884e-05,
+      "loss": 1.2962,
+      "step": 19832
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.0075230598449707,
+      "learning_rate": 4.475997885276236e-05,
+      "loss": 1.3067,
+      "step": 19833
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.745767593383789,
+      "learning_rate": 4.475971451229183e-05,
+      "loss": 1.2608,
+      "step": 19834
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8412258625030518,
+      "learning_rate": 4.475945017182131e-05,
+      "loss": 1.2854,
+      "step": 19835
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7781506776809692,
+      "learning_rate": 4.475918583135078e-05,
+      "loss": 1.25,
+      "step": 19836
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6950637102127075,
+      "learning_rate": 4.475892149088026e-05,
+      "loss": 1.3609,
+      "step": 19837
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9803920984268188,
+      "learning_rate": 4.4758657150409725e-05,
+      "loss": 1.3683,
+      "step": 19838
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9009746313095093,
+      "learning_rate": 4.4758392809939206e-05,
+      "loss": 1.2155,
+      "step": 19839
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2912471294403076,
+      "learning_rate": 4.475812846946868e-05,
+      "loss": 1.2237,
+      "step": 19840
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0730814933776855,
+      "learning_rate": 4.475786412899815e-05,
+      "loss": 2.1608,
+      "step": 19841
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9370356798171997,
+      "learning_rate": 4.4757599788527627e-05,
+      "loss": 1.2396,
+      "step": 19842
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.146786689758301,
+      "learning_rate": 4.47573354480571e-05,
+      "loss": 1.4606,
+      "step": 19843
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4424198865890503,
+      "learning_rate": 4.4757071107586574e-05,
+      "loss": 1.1813,
+      "step": 19844
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8576414585113525,
+      "learning_rate": 4.475680676711605e-05,
+      "loss": 2.4085,
+      "step": 19845
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.016188859939575,
+      "learning_rate": 4.475654242664552e-05,
+      "loss": 1.3276,
+      "step": 19846
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.777134895324707,
+      "learning_rate": 4.4756278086174994e-05,
+      "loss": 1.3954,
+      "step": 19847
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9578328132629395,
+      "learning_rate": 4.475601374570447e-05,
+      "loss": 1.2588,
+      "step": 19848
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0497758388519287,
+      "learning_rate": 4.475574940523394e-05,
+      "loss": 1.2579,
+      "step": 19849
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8444507122039795,
+      "learning_rate": 4.475548506476342e-05,
+      "loss": 1.3806,
+      "step": 19850
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8893872499465942,
+      "learning_rate": 4.475522072429289e-05,
+      "loss": 2.1864,
+      "step": 19851
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7075430154800415,
+      "learning_rate": 4.475495638382237e-05,
+      "loss": 1.1988,
+      "step": 19852
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.148252010345459,
+      "learning_rate": 4.4754692043351836e-05,
+      "loss": 1.282,
+      "step": 19853
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9475438594818115,
+      "learning_rate": 4.4754427702881316e-05,
+      "loss": 1.1263,
+      "step": 19854
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6474533081054688,
+      "learning_rate": 4.475416336241079e-05,
+      "loss": 1.2504,
+      "step": 19855
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7418421506881714,
+      "learning_rate": 4.4753899021940257e-05,
+      "loss": 1.2742,
+      "step": 19856
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0675604343414307,
+      "learning_rate": 4.475363468146974e-05,
+      "loss": 1.456,
+      "step": 19857
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.827230453491211,
+      "learning_rate": 4.4753370340999204e-05,
+      "loss": 1.2614,
+      "step": 19858
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8298225402832031,
+      "learning_rate": 4.4753106000528684e-05,
+      "loss": 1.2849,
+      "step": 19859
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8021609783172607,
+      "learning_rate": 4.475284166005816e-05,
+      "loss": 1.3896,
+      "step": 19860
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9446369409561157,
+      "learning_rate": 4.475257731958763e-05,
+      "loss": 1.2299,
+      "step": 19861
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.5601823329925537,
+      "learning_rate": 4.4752312979117105e-05,
+      "loss": 1.2926,
+      "step": 19862
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.996530294418335,
+      "learning_rate": 4.475204863864658e-05,
+      "loss": 2.4042,
+      "step": 19863
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6637201309204102,
+      "learning_rate": 4.475178429817605e-05,
+      "loss": 1.2016,
+      "step": 19864
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.801530599594116,
+      "learning_rate": 4.475151995770553e-05,
+      "loss": 1.3497,
+      "step": 19865
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.178194999694824,
+      "learning_rate": 4.4751255617235e-05,
+      "loss": 1.2476,
+      "step": 19866
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7047661542892456,
+      "learning_rate": 4.475099127676447e-05,
+      "loss": 1.2138,
+      "step": 19867
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.3873831033706665,
+      "learning_rate": 4.4750726936293946e-05,
+      "loss": 1.1275,
+      "step": 19868
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.617606282234192,
+      "learning_rate": 4.475046259582342e-05,
+      "loss": 1.2556,
+      "step": 19869
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1924550533294678,
+      "learning_rate": 4.47501982553529e-05,
+      "loss": 1.3284,
+      "step": 19870
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1132020950317383,
+      "learning_rate": 4.474993391488237e-05,
+      "loss": 1.3169,
+      "step": 19871
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1285934448242188,
+      "learning_rate": 4.474966957441185e-05,
+      "loss": 1.2683,
+      "step": 19872
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6489603519439697,
+      "learning_rate": 4.4749405233941314e-05,
+      "loss": 1.2363,
+      "step": 19873
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.672650694847107,
+      "learning_rate": 4.4749140893470794e-05,
+      "loss": 1.1748,
+      "step": 19874
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.253382682800293,
+      "learning_rate": 4.474887655300027e-05,
+      "loss": 1.286,
+      "step": 19875
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4540746212005615,
+      "learning_rate": 4.474861221252974e-05,
+      "loss": 1.3189,
+      "step": 19876
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7030963897705078,
+      "learning_rate": 4.4748347872059215e-05,
+      "loss": 1.145,
+      "step": 19877
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.230099678039551,
+      "learning_rate": 4.474808353158869e-05,
+      "loss": 1.3723,
+      "step": 19878
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8958371877670288,
+      "learning_rate": 4.474781919111816e-05,
+      "loss": 1.3357,
+      "step": 19879
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.9630966186523438,
+      "learning_rate": 4.4747554850647636e-05,
+      "loss": 1.3593,
+      "step": 19880
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.3052899837493896,
+      "learning_rate": 4.474729051017711e-05,
+      "loss": 1.4578,
+      "step": 19881
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.5760449171066284,
+      "learning_rate": 4.474702616970658e-05,
+      "loss": 1.212,
+      "step": 19882
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7611427307128906,
+      "learning_rate": 4.4746761829236056e-05,
+      "loss": 1.2397,
+      "step": 19883
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.390244483947754,
+      "learning_rate": 4.474649748876553e-05,
+      "loss": 1.3134,
+      "step": 19884
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.73643958568573,
+      "learning_rate": 4.474623314829501e-05,
+      "loss": 1.3307,
+      "step": 19885
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.813159465789795,
+      "learning_rate": 4.474596880782448e-05,
+      "loss": 1.208,
+      "step": 19886
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.869093179702759,
+      "learning_rate": 4.474570446735396e-05,
+      "loss": 1.3479,
+      "step": 19887
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0434842109680176,
+      "learning_rate": 4.4745440126883424e-05,
+      "loss": 1.3928,
+      "step": 19888
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.770888090133667,
+      "learning_rate": 4.4745175786412905e-05,
+      "loss": 1.3678,
+      "step": 19889
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.779447078704834,
+      "learning_rate": 4.474491144594238e-05,
+      "loss": 1.3163,
+      "step": 19890
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8072049617767334,
+      "learning_rate": 4.474464710547185e-05,
+      "loss": 1.1753,
+      "step": 19891
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.162775993347168,
+      "learning_rate": 4.4744382765001325e-05,
+      "loss": 1.2694,
+      "step": 19892
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2960948944091797,
+      "learning_rate": 4.474411842453079e-05,
+      "loss": 1.3414,
+      "step": 19893
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7656512260437012,
+      "learning_rate": 4.474385408406027e-05,
+      "loss": 1.1615,
+      "step": 19894
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0449795722961426,
+      "learning_rate": 4.4743589743589746e-05,
+      "loss": 1.3741,
+      "step": 19895
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4338841438293457,
+      "learning_rate": 4.474332540311922e-05,
+      "loss": 1.6068,
+      "step": 19896
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.3991061449050903,
+      "learning_rate": 4.474306106264869e-05,
+      "loss": 1.7284,
+      "step": 19897
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.893938660621643,
+      "learning_rate": 4.474279672217817e-05,
+      "loss": 1.3533,
+      "step": 19898
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.3334860801696777,
+      "learning_rate": 4.474253238170764e-05,
+      "loss": 1.4603,
+      "step": 19899
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.038325548171997,
+      "learning_rate": 4.474226804123712e-05,
+      "loss": 1.2407,
+      "step": 19900
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.245518445968628,
+      "learning_rate": 4.474200370076659e-05,
+      "loss": 1.3634,
+      "step": 19901
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9779078960418701,
+      "learning_rate": 4.474173936029606e-05,
+      "loss": 1.262,
+      "step": 19902
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.8242604732513428,
+      "learning_rate": 4.4741475019825535e-05,
+      "loss": 1.3521,
+      "step": 19903
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.316357374191284,
+      "learning_rate": 4.474121067935501e-05,
+      "loss": 1.4817,
+      "step": 19904
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0415918827056885,
+      "learning_rate": 4.474094633888449e-05,
+      "loss": 1.2794,
+      "step": 19905
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6189055442810059,
+      "learning_rate": 4.4740681998413955e-05,
+      "loss": 1.3384,
+      "step": 19906
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0950098037719727,
+      "learning_rate": 4.4740417657943436e-05,
+      "loss": 1.3935,
+      "step": 19907
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.765247106552124,
+      "learning_rate": 4.47401533174729e-05,
+      "loss": 1.2257,
+      "step": 19908
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8489100933074951,
+      "learning_rate": 4.473988897700238e-05,
+      "loss": 1.3227,
+      "step": 19909
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.772864818572998,
+      "learning_rate": 4.4739624636531856e-05,
+      "loss": 1.8054,
+      "step": 19910
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.3401954174041748,
+      "learning_rate": 4.473936029606133e-05,
+      "loss": 1.1361,
+      "step": 19911
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.129946708679199,
+      "learning_rate": 4.4739095955590804e-05,
+      "loss": 1.499,
+      "step": 19912
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.5213940143585205,
+      "learning_rate": 4.473883161512028e-05,
+      "loss": 1.277,
+      "step": 19913
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.388514995574951,
+      "learning_rate": 4.473856727464975e-05,
+      "loss": 1.2287,
+      "step": 19914
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9073257446289062,
+      "learning_rate": 4.4738302934179224e-05,
+      "loss": 1.3083,
+      "step": 19915
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.867708683013916,
+      "learning_rate": 4.47380385937087e-05,
+      "loss": 1.1518,
+      "step": 19916
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.040259599685669,
+      "learning_rate": 4.473777425323817e-05,
+      "loss": 1.2993,
+      "step": 19917
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.955791473388672,
+      "learning_rate": 4.4737509912767645e-05,
+      "loss": 1.2958,
+      "step": 19918
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.4866909980773926,
+      "learning_rate": 4.473724557229712e-05,
+      "loss": 1.2648,
+      "step": 19919
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.964609980583191,
+      "learning_rate": 4.47369812318266e-05,
+      "loss": 1.3701,
+      "step": 19920
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.6228084564208984,
+      "learning_rate": 4.4736716891356066e-05,
+      "loss": 1.329,
+      "step": 19921
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.3874611854553223,
+      "learning_rate": 4.4736452550885546e-05,
+      "loss": 1.3928,
+      "step": 19922
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0575180053710938,
+      "learning_rate": 4.473618821041501e-05,
+      "loss": 1.3646,
+      "step": 19923
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0297296047210693,
+      "learning_rate": 4.473592386994449e-05,
+      "loss": 1.3541,
+      "step": 19924
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6512490510940552,
+      "learning_rate": 4.473565952947397e-05,
+      "loss": 2.3592,
+      "step": 19925
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9415003061294556,
+      "learning_rate": 4.473539518900344e-05,
+      "loss": 1.3803,
+      "step": 19926
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9046001434326172,
+      "learning_rate": 4.4735130848532914e-05,
+      "loss": 1.1824,
+      "step": 19927
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9953255653381348,
+      "learning_rate": 4.473486650806239e-05,
+      "loss": 1.276,
+      "step": 19928
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.755183219909668,
+      "learning_rate": 4.473460216759186e-05,
+      "loss": 1.3091,
+      "step": 19929
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9618593454360962,
+      "learning_rate": 4.4734337827121335e-05,
+      "loss": 1.2332,
+      "step": 19930
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7888143062591553,
+      "learning_rate": 4.473407348665081e-05,
+      "loss": 1.1362,
+      "step": 19931
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8314478397369385,
+      "learning_rate": 4.473380914618028e-05,
+      "loss": 1.2153,
+      "step": 19932
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.006392002105713,
+      "learning_rate": 4.4733544805709755e-05,
+      "loss": 1.2556,
+      "step": 19933
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1088039875030518,
+      "learning_rate": 4.473328046523923e-05,
+      "loss": 1.3004,
+      "step": 19934
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2157912254333496,
+      "learning_rate": 4.473301612476871e-05,
+      "loss": 1.3811,
+      "step": 19935
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1837735176086426,
+      "learning_rate": 4.4732751784298176e-05,
+      "loss": 1.2989,
+      "step": 19936
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9023793935775757,
+      "learning_rate": 4.4732487443827656e-05,
+      "loss": 1.1794,
+      "step": 19937
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9645538330078125,
+      "learning_rate": 4.473222310335712e-05,
+      "loss": 1.3116,
+      "step": 19938
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1033308506011963,
+      "learning_rate": 4.47319587628866e-05,
+      "loss": 1.3663,
+      "step": 19939
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.962027907371521,
+      "learning_rate": 4.473169442241608e-05,
+      "loss": 1.2404,
+      "step": 19940
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2605977058410645,
+      "learning_rate": 4.4731430081945544e-05,
+      "loss": 1.3327,
+      "step": 19941
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8853235244750977,
+      "learning_rate": 4.4731165741475024e-05,
+      "loss": 1.3035,
+      "step": 19942
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.2973902225494385,
+      "learning_rate": 4.473090140100449e-05,
+      "loss": 1.4273,
+      "step": 19943
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.244619131088257,
+      "learning_rate": 4.473063706053397e-05,
+      "loss": 1.3283,
+      "step": 19944
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.140359878540039,
+      "learning_rate": 4.4730372720063445e-05,
+      "loss": 1.3879,
+      "step": 19945
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.064152240753174,
+      "learning_rate": 4.473010837959292e-05,
+      "loss": 1.2502,
+      "step": 19946
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8983523845672607,
+      "learning_rate": 4.472984403912239e-05,
+      "loss": 1.3492,
+      "step": 19947
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6549969911575317,
+      "learning_rate": 4.4729579698651866e-05,
+      "loss": 1.0766,
+      "step": 19948
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.957048773765564,
+      "learning_rate": 4.472931535818134e-05,
+      "loss": 1.2792,
+      "step": 19949
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2556145191192627,
+      "learning_rate": 4.472905101771081e-05,
+      "loss": 1.1692,
+      "step": 19950
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0296475887298584,
+      "learning_rate": 4.4728786677240286e-05,
+      "loss": 1.27,
+      "step": 19951
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.045156955718994,
+      "learning_rate": 4.472852233676976e-05,
+      "loss": 1.3141,
+      "step": 19952
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.5865259170532227,
+      "learning_rate": 4.4728257996299234e-05,
+      "loss": 1.0954,
+      "step": 19953
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.171405076980591,
+      "learning_rate": 4.472799365582871e-05,
+      "loss": 1.3329,
+      "step": 19954
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0725929737091064,
+      "learning_rate": 4.472772931535819e-05,
+      "loss": 1.2108,
+      "step": 19955
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.041416883468628,
+      "learning_rate": 4.4727464974887654e-05,
+      "loss": 1.3451,
+      "step": 19956
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.280980348587036,
+      "learning_rate": 4.4727200634417135e-05,
+      "loss": 1.3401,
+      "step": 19957
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.003810167312622,
+      "learning_rate": 4.47269362939466e-05,
+      "loss": 1.1264,
+      "step": 19958
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6093330383300781,
+      "learning_rate": 4.472667195347608e-05,
+      "loss": 1.2206,
+      "step": 19959
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.673842191696167,
+      "learning_rate": 4.4726407613005555e-05,
+      "loss": 1.2235,
+      "step": 19960
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.323235511779785,
+      "learning_rate": 4.472614327253503e-05,
+      "loss": 1.4846,
+      "step": 19961
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0850179195404053,
+      "learning_rate": 4.47258789320645e-05,
+      "loss": 1.345,
+      "step": 19962
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.407710313796997,
+      "learning_rate": 4.4725614591593976e-05,
+      "loss": 1.3035,
+      "step": 19963
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8046735525131226,
+      "learning_rate": 4.472535025112345e-05,
+      "loss": 1.2431,
+      "step": 19964
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.321718454360962,
+      "learning_rate": 4.472508591065292e-05,
+      "loss": 1.2744,
+      "step": 19965
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.14378023147583,
+      "learning_rate": 4.47248215701824e-05,
+      "loss": 1.4083,
+      "step": 19966
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.5215752124786377,
+      "learning_rate": 4.472455722971187e-05,
+      "loss": 1.4702,
+      "step": 19967
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2668190002441406,
+      "learning_rate": 4.4724292889241344e-05,
+      "loss": 1.2817,
+      "step": 19968
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0818986892700195,
+      "learning_rate": 4.472402854877082e-05,
+      "loss": 1.1542,
+      "step": 19969
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.870347261428833,
+      "learning_rate": 4.47237642083003e-05,
+      "loss": 1.9733,
+      "step": 19970
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8027981519699097,
+      "learning_rate": 4.4723499867829765e-05,
+      "loss": 1.3072,
+      "step": 19971
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1993165016174316,
+      "learning_rate": 4.4723235527359245e-05,
+      "loss": 1.2733,
+      "step": 19972
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.034005880355835,
+      "learning_rate": 4.472297118688871e-05,
+      "loss": 1.2518,
+      "step": 19973
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.5175459384918213,
+      "learning_rate": 4.4722706846418185e-05,
+      "loss": 1.3812,
+      "step": 19974
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1500484943389893,
+      "learning_rate": 4.4722442505947666e-05,
+      "loss": 1.3897,
+      "step": 19975
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2334187030792236,
+      "learning_rate": 4.472217816547713e-05,
+      "loss": 1.3859,
+      "step": 19976
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.065913200378418,
+      "learning_rate": 4.472191382500661e-05,
+      "loss": 1.2141,
+      "step": 19977
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.847751259803772,
+      "learning_rate": 4.472164948453608e-05,
+      "loss": 1.2355,
+      "step": 19978
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6705732345581055,
+      "learning_rate": 4.472138514406556e-05,
+      "loss": 1.2646,
+      "step": 19979
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0635340213775635,
+      "learning_rate": 4.4721120803595034e-05,
+      "loss": 1.2367,
+      "step": 19980
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.4141011238098145,
+      "learning_rate": 4.472085646312451e-05,
+      "loss": 1.4849,
+      "step": 19981
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.359916925430298,
+      "learning_rate": 4.472059212265398e-05,
+      "loss": 1.25,
+      "step": 19982
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.4498138427734375,
+      "learning_rate": 4.4720327782183454e-05,
+      "loss": 1.3384,
+      "step": 19983
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0306262969970703,
+      "learning_rate": 4.472006344171293e-05,
+      "loss": 1.345,
+      "step": 19984
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8758822679519653,
+      "learning_rate": 4.47197991012424e-05,
+      "loss": 1.3238,
+      "step": 19985
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.7147936820983887,
+      "learning_rate": 4.4719534760771875e-05,
+      "loss": 1.4581,
+      "step": 19986
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5281423330307007,
+      "learning_rate": 4.471927042030135e-05,
+      "loss": 1.0807,
+      "step": 19987
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.150081157684326,
+      "learning_rate": 4.471900607983082e-05,
+      "loss": 1.1342,
+      "step": 19988
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8985745906829834,
+      "learning_rate": 4.4718741739360296e-05,
+      "loss": 1.2414,
+      "step": 19989
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9964349269866943,
+      "learning_rate": 4.4718477398889776e-05,
+      "loss": 1.2945,
+      "step": 19990
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.85306715965271,
+      "learning_rate": 4.471821305841924e-05,
+      "loss": 1.2196,
+      "step": 19991
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.4460043907165527,
+      "learning_rate": 4.471794871794872e-05,
+      "loss": 1.4799,
+      "step": 19992
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.016859292984009,
+      "learning_rate": 4.471768437747819e-05,
+      "loss": 1.2492,
+      "step": 19993
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7195560932159424,
+      "learning_rate": 4.471742003700767e-05,
+      "loss": 1.261,
+      "step": 19994
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6726312637329102,
+      "learning_rate": 4.4717155696537144e-05,
+      "loss": 1.1974,
+      "step": 19995
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8866369724273682,
+      "learning_rate": 4.471689135606662e-05,
+      "loss": 1.3451,
+      "step": 19996
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1442503929138184,
+      "learning_rate": 4.471662701559609e-05,
+      "loss": 1.4191,
+      "step": 19997
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6675050258636475,
+      "learning_rate": 4.4716362675125565e-05,
+      "loss": 1.2914,
+      "step": 19998
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2635860443115234,
+      "learning_rate": 4.471609833465504e-05,
+      "loss": 1.5074,
+      "step": 19999
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9443665742874146,
+      "learning_rate": 4.471583399418451e-05,
+      "loss": 1.3155,
+      "step": 20000
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1080074310302734,
+      "learning_rate": 4.4715569653713985e-05,
+      "loss": 1.4263,
+      "step": 20001
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7205288410186768,
+      "learning_rate": 4.471530531324346e-05,
+      "loss": 1.2744,
+      "step": 20002
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0063443183898926,
+      "learning_rate": 4.471504097277293e-05,
+      "loss": 1.4031,
+      "step": 20003
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6252756118774414,
+      "learning_rate": 4.4714776632302406e-05,
+      "loss": 1.0569,
+      "step": 20004
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.995240330696106,
+      "learning_rate": 4.4714512291831886e-05,
+      "loss": 1.2908,
+      "step": 20005
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6722936630249023,
+      "learning_rate": 4.471424795136135e-05,
+      "loss": 1.0613,
+      "step": 20006
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2738497257232666,
+      "learning_rate": 4.4713983610890833e-05,
+      "loss": 1.3903,
+      "step": 20007
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8854719400405884,
+      "learning_rate": 4.47137192704203e-05,
+      "loss": 1.2495,
+      "step": 20008
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.047926425933838,
+      "learning_rate": 4.471345492994978e-05,
+      "loss": 1.3882,
+      "step": 20009
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.610993504524231,
+      "learning_rate": 4.4713190589479254e-05,
+      "loss": 1.2552,
+      "step": 20010
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.618227481842041,
+      "learning_rate": 4.471292624900872e-05,
+      "loss": 1.1977,
+      "step": 20011
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8168280124664307,
+      "learning_rate": 4.47126619085382e-05,
+      "loss": 1.3143,
+      "step": 20012
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9892830848693848,
+      "learning_rate": 4.471239756806767e-05,
+      "loss": 1.2101,
+      "step": 20013
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.469224214553833,
+      "learning_rate": 4.471213322759715e-05,
+      "loss": 1.9389,
+      "step": 20014
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9097466468811035,
+      "learning_rate": 4.471186888712662e-05,
+      "loss": 1.2643,
+      "step": 20015
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5873368978500366,
+      "learning_rate": 4.4711604546656096e-05,
+      "loss": 1.1767,
+      "step": 20016
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8964468240737915,
+      "learning_rate": 4.471134020618557e-05,
+      "loss": 1.454,
+      "step": 20017
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8400593996047974,
+      "learning_rate": 4.471107586571504e-05,
+      "loss": 1.3447,
+      "step": 20018
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7639082670211792,
+      "learning_rate": 4.4710811525244516e-05,
+      "loss": 1.228,
+      "step": 20019
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.8250815868377686,
+      "learning_rate": 4.471054718477399e-05,
+      "loss": 1.459,
+      "step": 20020
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.4084348678588867,
+      "learning_rate": 4.4710282844303463e-05,
+      "loss": 1.4014,
+      "step": 20021
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7471320629119873,
+      "learning_rate": 4.471001850383294e-05,
+      "loss": 1.3496,
+      "step": 20022
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.973347783088684,
+      "learning_rate": 4.470975416336241e-05,
+      "loss": 1.2411,
+      "step": 20023
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7325153350830078,
+      "learning_rate": 4.4709489822891884e-05,
+      "loss": 1.1395,
+      "step": 20024
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7420209646224976,
+      "learning_rate": 4.4709225482421365e-05,
+      "loss": 1.2281,
+      "step": 20025
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9848418235778809,
+      "learning_rate": 4.470896114195083e-05,
+      "loss": 1.3363,
+      "step": 20026
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6992801427841187,
+      "learning_rate": 4.470869680148031e-05,
+      "loss": 1.6679,
+      "step": 20027
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0148184299468994,
+      "learning_rate": 4.470843246100978e-05,
+      "loss": 1.3009,
+      "step": 20028
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.02115797996521,
+      "learning_rate": 4.470816812053926e-05,
+      "loss": 1.4149,
+      "step": 20029
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6488749980926514,
+      "learning_rate": 4.470790378006873e-05,
+      "loss": 1.2015,
+      "step": 20030
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2919631004333496,
+      "learning_rate": 4.4707639439598206e-05,
+      "loss": 1.2695,
+      "step": 20031
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.924676537513733,
+      "learning_rate": 4.470737509912768e-05,
+      "loss": 1.4836,
+      "step": 20032
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.6377978324890137,
+      "learning_rate": 4.470711075865715e-05,
+      "loss": 1.4966,
+      "step": 20033
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8409569263458252,
+      "learning_rate": 4.470684641818663e-05,
+      "loss": 1.3279,
+      "step": 20034
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.4000697135925293,
+      "learning_rate": 4.47065820777161e-05,
+      "loss": 1.4125,
+      "step": 20035
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6868908405303955,
+      "learning_rate": 4.4706317737245574e-05,
+      "loss": 1.0243,
+      "step": 20036
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.7222647666931152,
+      "learning_rate": 4.470605339677505e-05,
+      "loss": 1.271,
+      "step": 20037
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.103018045425415,
+      "learning_rate": 4.470578905630452e-05,
+      "loss": 1.2319,
+      "step": 20038
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8848062753677368,
+      "learning_rate": 4.4705524715833995e-05,
+      "loss": 1.2979,
+      "step": 20039
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.3066673278808594,
+      "learning_rate": 4.4705260375363475e-05,
+      "loss": 1.5147,
+      "step": 20040
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6980886459350586,
+      "learning_rate": 4.470499603489294e-05,
+      "loss": 2.1933,
+      "step": 20041
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.016832113265991,
+      "learning_rate": 4.470473169442242e-05,
+      "loss": 1.3867,
+      "step": 20042
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9770362377166748,
+      "learning_rate": 4.470446735395189e-05,
+      "loss": 1.9321,
+      "step": 20043
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7942497730255127,
+      "learning_rate": 4.470420301348137e-05,
+      "loss": 1.1614,
+      "step": 20044
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9509644508361816,
+      "learning_rate": 4.470393867301084e-05,
+      "loss": 1.2408,
+      "step": 20045
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8067582845687866,
+      "learning_rate": 4.470367433254031e-05,
+      "loss": 1.1985,
+      "step": 20046
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2228100299835205,
+      "learning_rate": 4.470340999206979e-05,
+      "loss": 1.384,
+      "step": 20047
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9284945726394653,
+      "learning_rate": 4.470314565159926e-05,
+      "loss": 1.2026,
+      "step": 20048
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2575979232788086,
+      "learning_rate": 4.470288131112874e-05,
+      "loss": 1.2819,
+      "step": 20049
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.726135492324829,
+      "learning_rate": 4.470261697065821e-05,
+      "loss": 1.1945,
+      "step": 20050
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.4058871269226074,
+      "learning_rate": 4.4702352630187684e-05,
+      "loss": 1.3122,
+      "step": 20051
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5963364839553833,
+      "learning_rate": 4.470208828971716e-05,
+      "loss": 1.1771,
+      "step": 20052
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5256599187850952,
+      "learning_rate": 4.470182394924663e-05,
+      "loss": 2.1145,
+      "step": 20053
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7382344007492065,
+      "learning_rate": 4.4701559608776105e-05,
+      "loss": 1.2477,
+      "step": 20054
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1997196674346924,
+      "learning_rate": 4.4701295268305585e-05,
+      "loss": 1.3962,
+      "step": 20055
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.3079540729522705,
+      "learning_rate": 4.470103092783505e-05,
+      "loss": 1.2098,
+      "step": 20056
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9297975301742554,
+      "learning_rate": 4.4700766587364526e-05,
+      "loss": 1.2882,
+      "step": 20057
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8771347999572754,
+      "learning_rate": 4.4700502246894e-05,
+      "loss": 1.3306,
+      "step": 20058
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7896010875701904,
+      "learning_rate": 4.470023790642347e-05,
+      "loss": 1.1997,
+      "step": 20059
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.92877197265625,
+      "learning_rate": 4.469997356595295e-05,
+      "loss": 1.3424,
+      "step": 20060
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2680084705352783,
+      "learning_rate": 4.469970922548242e-05,
+      "loss": 1.2551,
+      "step": 20061
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.168227195739746,
+      "learning_rate": 4.46994448850119e-05,
+      "loss": 1.3306,
+      "step": 20062
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.682181715965271,
+      "learning_rate": 4.469918054454137e-05,
+      "loss": 1.5305,
+      "step": 20063
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.94923734664917,
+      "learning_rate": 4.469891620407085e-05,
+      "loss": 1.4924,
+      "step": 20064
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.101027011871338,
+      "learning_rate": 4.469865186360032e-05,
+      "loss": 1.3547,
+      "step": 20065
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9067941904067993,
+      "learning_rate": 4.4698387523129794e-05,
+      "loss": 1.2999,
+      "step": 20066
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7794402837753296,
+      "learning_rate": 4.469812318265927e-05,
+      "loss": 1.2577,
+      "step": 20067
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.403974771499634,
+      "learning_rate": 4.469785884218874e-05,
+      "loss": 1.2884,
+      "step": 20068
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9085702896118164,
+      "learning_rate": 4.4697594501718215e-05,
+      "loss": 1.2877,
+      "step": 20069
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.887559413909912,
+      "learning_rate": 4.469733016124769e-05,
+      "loss": 1.3316,
+      "step": 20070
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.721587061882019,
+      "learning_rate": 4.469706582077716e-05,
+      "loss": 2.2552,
+      "step": 20071
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.313178777694702,
+      "learning_rate": 4.4696801480306636e-05,
+      "loss": 1.3386,
+      "step": 20072
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1280133724212646,
+      "learning_rate": 4.469653713983611e-05,
+      "loss": 1.3094,
+      "step": 20073
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2881102561950684,
+      "learning_rate": 4.469627279936558e-05,
+      "loss": 1.3012,
+      "step": 20074
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0647172927856445,
+      "learning_rate": 4.4696008458895063e-05,
+      "loss": 1.2515,
+      "step": 20075
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7346261739730835,
+      "learning_rate": 4.469574411842453e-05,
+      "loss": 1.3048,
+      "step": 20076
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.114121675491333,
+      "learning_rate": 4.469547977795401e-05,
+      "loss": 1.3645,
+      "step": 20077
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.247610092163086,
+      "learning_rate": 4.469521543748348e-05,
+      "loss": 1.4097,
+      "step": 20078
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7927558422088623,
+      "learning_rate": 4.469495109701296e-05,
+      "loss": 1.3683,
+      "step": 20079
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.889549732208252,
+      "learning_rate": 4.469468675654243e-05,
+      "loss": 1.291,
+      "step": 20080
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7111436128616333,
+      "learning_rate": 4.4694422416071905e-05,
+      "loss": 1.1187,
+      "step": 20081
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.277489185333252,
+      "learning_rate": 4.469415807560138e-05,
+      "loss": 1.2332,
+      "step": 20082
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.310659646987915,
+      "learning_rate": 4.4693893735130845e-05,
+      "loss": 1.3776,
+      "step": 20083
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.045565366744995,
+      "learning_rate": 4.4693629394660326e-05,
+      "loss": 1.2114,
+      "step": 20084
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.795464038848877,
+      "learning_rate": 4.46933650541898e-05,
+      "loss": 1.3278,
+      "step": 20085
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7030023336410522,
+      "learning_rate": 4.469310071371927e-05,
+      "loss": 1.3204,
+      "step": 20086
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6922286748886108,
+      "learning_rate": 4.4692836373248746e-05,
+      "loss": 1.2097,
+      "step": 20087
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6992228031158447,
+      "learning_rate": 4.469257203277822e-05,
+      "loss": 1.2246,
+      "step": 20088
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9323291778564453,
+      "learning_rate": 4.4692307692307693e-05,
+      "loss": 1.4428,
+      "step": 20089
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.4630846977233887,
+      "learning_rate": 4.4692043351837174e-05,
+      "loss": 1.1984,
+      "step": 20090
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.970404028892517,
+      "learning_rate": 4.469177901136664e-05,
+      "loss": 1.343,
+      "step": 20091
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0257680416107178,
+      "learning_rate": 4.4691514670896114e-05,
+      "loss": 1.3959,
+      "step": 20092
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9909511804580688,
+      "learning_rate": 4.469125033042559e-05,
+      "loss": 1.3132,
+      "step": 20093
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.390789747238159,
+      "learning_rate": 4.469098598995506e-05,
+      "loss": 1.4224,
+      "step": 20094
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.116001605987549,
+      "learning_rate": 4.469072164948454e-05,
+      "loss": 1.1959,
+      "step": 20095
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.5052735805511475,
+      "learning_rate": 4.469045730901401e-05,
+      "loss": 1.2627,
+      "step": 20096
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8088247776031494,
+      "learning_rate": 4.469019296854349e-05,
+      "loss": 1.2491,
+      "step": 20097
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6795727014541626,
+      "learning_rate": 4.4689928628072956e-05,
+      "loss": 1.1889,
+      "step": 20098
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.442272663116455,
+      "learning_rate": 4.4689664287602436e-05,
+      "loss": 1.3466,
+      "step": 20099
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.968729257583618,
+      "learning_rate": 4.468939994713191e-05,
+      "loss": 1.4262,
+      "step": 20100
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.132812261581421,
+      "learning_rate": 4.468913560666138e-05,
+      "loss": 1.3181,
+      "step": 20101
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.3730437755584717,
+      "learning_rate": 4.4688871266190857e-05,
+      "loss": 1.3655,
+      "step": 20102
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.043164014816284,
+      "learning_rate": 4.468860692572033e-05,
+      "loss": 1.2286,
+      "step": 20103
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8806761503219604,
+      "learning_rate": 4.4688342585249804e-05,
+      "loss": 1.2684,
+      "step": 20104
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0361196994781494,
+      "learning_rate": 4.468807824477928e-05,
+      "loss": 1.3841,
+      "step": 20105
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.274061679840088,
+      "learning_rate": 4.468781390430875e-05,
+      "loss": 1.2972,
+      "step": 20106
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7855805158615112,
+      "learning_rate": 4.4687549563838224e-05,
+      "loss": 1.3341,
+      "step": 20107
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.157632827758789,
+      "learning_rate": 4.46872852233677e-05,
+      "loss": 1.2505,
+      "step": 20108
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.3005177974700928,
+      "learning_rate": 4.468702088289717e-05,
+      "loss": 1.1809,
+      "step": 20109
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.138186454772949,
+      "learning_rate": 4.468675654242665e-05,
+      "loss": 1.3533,
+      "step": 20110
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0023605823516846,
+      "learning_rate": 4.468649220195612e-05,
+      "loss": 1.2788,
+      "step": 20111
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6920490264892578,
+      "learning_rate": 4.46862278614856e-05,
+      "loss": 1.2428,
+      "step": 20112
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7654060125350952,
+      "learning_rate": 4.4685963521015066e-05,
+      "loss": 1.396,
+      "step": 20113
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.4036550521850586,
+      "learning_rate": 4.4685699180544546e-05,
+      "loss": 1.2819,
+      "step": 20114
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.899546504020691,
+      "learning_rate": 4.468543484007402e-05,
+      "loss": 1.2928,
+      "step": 20115
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.103445291519165,
+      "learning_rate": 4.468517049960349e-05,
+      "loss": 1.4045,
+      "step": 20116
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.6886227130889893,
+      "learning_rate": 4.468490615913297e-05,
+      "loss": 1.3928,
+      "step": 20117
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7954922914505005,
+      "learning_rate": 4.468464181866244e-05,
+      "loss": 1.2245,
+      "step": 20118
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2478859424591064,
+      "learning_rate": 4.4684377478191914e-05,
+      "loss": 1.3451,
+      "step": 20119
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6443346738815308,
+      "learning_rate": 4.468411313772139e-05,
+      "loss": 1.404,
+      "step": 20120
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.349809169769287,
+      "learning_rate": 4.468384879725086e-05,
+      "loss": 1.4259,
+      "step": 20121
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6046042442321777,
+      "learning_rate": 4.4683584456780335e-05,
+      "loss": 1.2762,
+      "step": 20122
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.541652679443359,
+      "learning_rate": 4.468332011630981e-05,
+      "loss": 1.1914,
+      "step": 20123
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5053799152374268,
+      "learning_rate": 4.468305577583928e-05,
+      "loss": 1.2153,
+      "step": 20124
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7848061323165894,
+      "learning_rate": 4.468279143536876e-05,
+      "loss": 1.1958,
+      "step": 20125
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7739708423614502,
+      "learning_rate": 4.468252709489823e-05,
+      "loss": 1.2254,
+      "step": 20126
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9295754432678223,
+      "learning_rate": 4.468226275442771e-05,
+      "loss": 1.4061,
+      "step": 20127
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9930706024169922,
+      "learning_rate": 4.4681998413957176e-05,
+      "loss": 1.2025,
+      "step": 20128
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.850218653678894,
+      "learning_rate": 4.468173407348665e-05,
+      "loss": 1.2936,
+      "step": 20129
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.989630937576294,
+      "learning_rate": 4.468146973301613e-05,
+      "loss": 1.6398,
+      "step": 20130
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6400530338287354,
+      "learning_rate": 4.46812053925456e-05,
+      "loss": 1.2297,
+      "step": 20131
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.4573583602905273,
+      "learning_rate": 4.468094105207508e-05,
+      "loss": 1.3781,
+      "step": 20132
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7553162574768066,
+      "learning_rate": 4.4680676711604544e-05,
+      "loss": 1.318,
+      "step": 20133
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8211946487426758,
+      "learning_rate": 4.4680412371134024e-05,
+      "loss": 1.2507,
+      "step": 20134
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.692874550819397,
+      "learning_rate": 4.46801480306635e-05,
+      "loss": 1.2029,
+      "step": 20135
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2329843044281006,
+      "learning_rate": 4.467988369019297e-05,
+      "loss": 1.2205,
+      "step": 20136
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.842773675918579,
+      "learning_rate": 4.4679619349722445e-05,
+      "loss": 1.4365,
+      "step": 20137
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6633758544921875,
+      "learning_rate": 4.467935500925192e-05,
+      "loss": 1.2394,
+      "step": 20138
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1845028400421143,
+      "learning_rate": 4.467909066878139e-05,
+      "loss": 1.436,
+      "step": 20139
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.064716339111328,
+      "learning_rate": 4.4678826328310866e-05,
+      "loss": 1.2699,
+      "step": 20140
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6366899013519287,
+      "learning_rate": 4.467856198784034e-05,
+      "loss": 1.2356,
+      "step": 20141
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9551146030426025,
+      "learning_rate": 4.467829764736981e-05,
+      "loss": 1.2979,
+      "step": 20142
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6107475757598877,
+      "learning_rate": 4.4678033306899287e-05,
+      "loss": 1.1872,
+      "step": 20143
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2917768955230713,
+      "learning_rate": 4.467776896642876e-05,
+      "loss": 1.45,
+      "step": 20144
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.5720791816711426,
+      "learning_rate": 4.467750462595824e-05,
+      "loss": 1.3174,
+      "step": 20145
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.4833946228027344,
+      "learning_rate": 4.467724028548771e-05,
+      "loss": 1.2666,
+      "step": 20146
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.940980076789856,
+      "learning_rate": 4.467697594501719e-05,
+      "loss": 1.1812,
+      "step": 20147
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8307945728302002,
+      "learning_rate": 4.4676711604546654e-05,
+      "loss": 1.2601,
+      "step": 20148
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9810229539871216,
+      "learning_rate": 4.4676447264076135e-05,
+      "loss": 1.3383,
+      "step": 20149
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6485072374343872,
+      "learning_rate": 4.467618292360561e-05,
+      "loss": 1.0961,
+      "step": 20150
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8811060190200806,
+      "learning_rate": 4.467591858313508e-05,
+      "loss": 1.2105,
+      "step": 20151
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.628261923789978,
+      "learning_rate": 4.4675654242664555e-05,
+      "loss": 1.0969,
+      "step": 20152
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9254491329193115,
+      "learning_rate": 4.467538990219403e-05,
+      "loss": 1.3267,
+      "step": 20153
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2720210552215576,
+      "learning_rate": 4.46751255617235e-05,
+      "loss": 2.0279,
+      "step": 20154
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.601426601409912,
+      "learning_rate": 4.4674861221252976e-05,
+      "loss": 1.3024,
+      "step": 20155
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2791621685028076,
+      "learning_rate": 4.467459688078245e-05,
+      "loss": 1.3682,
+      "step": 20156
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6247437000274658,
+      "learning_rate": 4.467433254031192e-05,
+      "loss": 1.121,
+      "step": 20157
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6623485088348389,
+      "learning_rate": 4.46740681998414e-05,
+      "loss": 1.9442,
+      "step": 20158
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8851122856140137,
+      "learning_rate": 4.467380385937087e-05,
+      "loss": 1.2809,
+      "step": 20159
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6457751989364624,
+      "learning_rate": 4.467353951890035e-05,
+      "loss": 1.288,
+      "step": 20160
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.122286081314087,
+      "learning_rate": 4.467327517842982e-05,
+      "loss": 1.3882,
+      "step": 20161
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.074406147003174,
+      "learning_rate": 4.46730108379593e-05,
+      "loss": 1.4336,
+      "step": 20162
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9192417860031128,
+      "learning_rate": 4.4672746497488765e-05,
+      "loss": 1.2711,
+      "step": 20163
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.080947160720825,
+      "learning_rate": 4.467248215701824e-05,
+      "loss": 1.3442,
+      "step": 20164
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5586705207824707,
+      "learning_rate": 4.467221781654772e-05,
+      "loss": 1.1334,
+      "step": 20165
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.629757881164551,
+      "learning_rate": 4.4671953476077185e-05,
+      "loss": 1.2871,
+      "step": 20166
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.77102792263031,
+      "learning_rate": 4.4671689135606666e-05,
+      "loss": 1.274,
+      "step": 20167
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.31782603263855,
+      "learning_rate": 4.467142479513613e-05,
+      "loss": 1.1978,
+      "step": 20168
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1690127849578857,
+      "learning_rate": 4.467116045466561e-05,
+      "loss": 1.4761,
+      "step": 20169
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1007704734802246,
+      "learning_rate": 4.4670896114195087e-05,
+      "loss": 1.3082,
+      "step": 20170
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.353586196899414,
+      "learning_rate": 4.467063177372456e-05,
+      "loss": 1.3794,
+      "step": 20171
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.716757893562317,
+      "learning_rate": 4.4670367433254034e-05,
+      "loss": 1.1507,
+      "step": 20172
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9513967037200928,
+      "learning_rate": 4.467010309278351e-05,
+      "loss": 1.3183,
+      "step": 20173
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1008248329162598,
+      "learning_rate": 4.466983875231298e-05,
+      "loss": 1.2556,
+      "step": 20174
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.3360300064086914,
+      "learning_rate": 4.4669574411842454e-05,
+      "loss": 1.3627,
+      "step": 20175
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9078010320663452,
+      "learning_rate": 4.466931007137193e-05,
+      "loss": 1.3753,
+      "step": 20176
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.825746774673462,
+      "learning_rate": 4.46690457309014e-05,
+      "loss": 1.2663,
+      "step": 20177
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4866280555725098,
+      "learning_rate": 4.4668781390430875e-05,
+      "loss": 1.2519,
+      "step": 20178
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6787060499191284,
+      "learning_rate": 4.466851704996035e-05,
+      "loss": 1.1282,
+      "step": 20179
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0591628551483154,
+      "learning_rate": 4.466825270948983e-05,
+      "loss": 1.3527,
+      "step": 20180
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.317294120788574,
+      "learning_rate": 4.4667988369019296e-05,
+      "loss": 1.375,
+      "step": 20181
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7846366167068481,
+      "learning_rate": 4.4667724028548776e-05,
+      "loss": 1.3344,
+      "step": 20182
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7252073287963867,
+      "learning_rate": 4.466745968807824e-05,
+      "loss": 1.918,
+      "step": 20183
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6066795587539673,
+      "learning_rate": 4.466719534760772e-05,
+      "loss": 1.2002,
+      "step": 20184
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9738601446151733,
+      "learning_rate": 4.46669310071372e-05,
+      "loss": 1.2983,
+      "step": 20185
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8361579179763794,
+      "learning_rate": 4.466666666666667e-05,
+      "loss": 1.2777,
+      "step": 20186
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4161477088928223,
+      "learning_rate": 4.4666402326196144e-05,
+      "loss": 1.3634,
+      "step": 20187
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.421719551086426,
+      "learning_rate": 4.466613798572562e-05,
+      "loss": 1.5237,
+      "step": 20188
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8447452783584595,
+      "learning_rate": 4.466587364525509e-05,
+      "loss": 1.7021,
+      "step": 20189
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.584078788757324,
+      "learning_rate": 4.4665609304784565e-05,
+      "loss": 1.4492,
+      "step": 20190
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4175171852111816,
+      "learning_rate": 4.466534496431404e-05,
+      "loss": 1.46,
+      "step": 20191
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4251937866210938,
+      "learning_rate": 4.466508062384351e-05,
+      "loss": 1.2239,
+      "step": 20192
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8178664445877075,
+      "learning_rate": 4.4664816283372985e-05,
+      "loss": 1.1268,
+      "step": 20193
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5852128267288208,
+      "learning_rate": 4.466455194290246e-05,
+      "loss": 1.2525,
+      "step": 20194
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7770438194274902,
+      "learning_rate": 4.466428760243194e-05,
+      "loss": 1.2397,
+      "step": 20195
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7606894969940186,
+      "learning_rate": 4.4664023261961406e-05,
+      "loss": 1.244,
+      "step": 20196
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.8912699222564697,
+      "learning_rate": 4.4663758921490886e-05,
+      "loss": 1.3417,
+      "step": 20197
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6577296257019043,
+      "learning_rate": 4.466349458102035e-05,
+      "loss": 1.1171,
+      "step": 20198
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.397677421569824,
+      "learning_rate": 4.4663230240549834e-05,
+      "loss": 1.38,
+      "step": 20199
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9758946895599365,
+      "learning_rate": 4.466296590007931e-05,
+      "loss": 1.4697,
+      "step": 20200
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.078627109527588,
+      "learning_rate": 4.4662701559608774e-05,
+      "loss": 1.2851,
+      "step": 20201
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.015364408493042,
+      "learning_rate": 4.4662437219138254e-05,
+      "loss": 1.3336,
+      "step": 20202
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2466771602630615,
+      "learning_rate": 4.466217287866772e-05,
+      "loss": 1.34,
+      "step": 20203
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8665215969085693,
+      "learning_rate": 4.46619085381972e-05,
+      "loss": 1.3556,
+      "step": 20204
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9105523824691772,
+      "learning_rate": 4.4661644197726675e-05,
+      "loss": 1.2831,
+      "step": 20205
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1645989418029785,
+      "learning_rate": 4.466137985725615e-05,
+      "loss": 1.2846,
+      "step": 20206
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8925657272338867,
+      "learning_rate": 4.466111551678562e-05,
+      "loss": 1.3304,
+      "step": 20207
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.820967435836792,
+      "learning_rate": 4.4660851176315096e-05,
+      "loss": 1.1267,
+      "step": 20208
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.870126485824585,
+      "learning_rate": 4.466058683584457e-05,
+      "loss": 1.2032,
+      "step": 20209
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1733124256134033,
+      "learning_rate": 4.466032249537404e-05,
+      "loss": 1.2148,
+      "step": 20210
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9323606491088867,
+      "learning_rate": 4.4660058154903516e-05,
+      "loss": 1.3078,
+      "step": 20211
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7903727293014526,
+      "learning_rate": 4.465979381443299e-05,
+      "loss": 1.3653,
+      "step": 20212
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.103710651397705,
+      "learning_rate": 4.4659529473962464e-05,
+      "loss": 1.3145,
+      "step": 20213
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0341265201568604,
+      "learning_rate": 4.465926513349194e-05,
+      "loss": 1.2312,
+      "step": 20214
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9552196264266968,
+      "learning_rate": 4.465900079302142e-05,
+      "loss": 1.2021,
+      "step": 20215
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.813732385635376,
+      "learning_rate": 4.4658736452550884e-05,
+      "loss": 1.3353,
+      "step": 20216
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.158881187438965,
+      "learning_rate": 4.4658472112080365e-05,
+      "loss": 1.6027,
+      "step": 20217
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.828359365463257,
+      "learning_rate": 4.465820777160983e-05,
+      "loss": 1.4273,
+      "step": 20218
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8129521608352661,
+      "learning_rate": 4.465794343113931e-05,
+      "loss": 1.1766,
+      "step": 20219
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.5377867221832275,
+      "learning_rate": 4.4657679090668785e-05,
+      "loss": 1.304,
+      "step": 20220
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8973767757415771,
+      "learning_rate": 4.465741475019826e-05,
+      "loss": 1.2421,
+      "step": 20221
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.349700689315796,
+      "learning_rate": 4.465715040972773e-05,
+      "loss": 1.3928,
+      "step": 20222
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4191040992736816,
+      "learning_rate": 4.4656886069257206e-05,
+      "loss": 1.3325,
+      "step": 20223
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4320895671844482,
+      "learning_rate": 4.465662172878668e-05,
+      "loss": 1.3091,
+      "step": 20224
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.615617036819458,
+      "learning_rate": 4.465635738831615e-05,
+      "loss": 1.2641,
+      "step": 20225
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.807649850845337,
+      "learning_rate": 4.465609304784563e-05,
+      "loss": 1.2781,
+      "step": 20226
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5540614128112793,
+      "learning_rate": 4.46558287073751e-05,
+      "loss": 1.1531,
+      "step": 20227
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.212644577026367,
+      "learning_rate": 4.4655564366904574e-05,
+      "loss": 1.4938,
+      "step": 20228
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.678896427154541,
+      "learning_rate": 4.465530002643405e-05,
+      "loss": 1.1978,
+      "step": 20229
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0072224140167236,
+      "learning_rate": 4.465503568596353e-05,
+      "loss": 1.5413,
+      "step": 20230
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2814037799835205,
+      "learning_rate": 4.4654771345492995e-05,
+      "loss": 1.2508,
+      "step": 20231
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0360803604125977,
+      "learning_rate": 4.4654507005022475e-05,
+      "loss": 1.1784,
+      "step": 20232
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.295835018157959,
+      "learning_rate": 4.465424266455194e-05,
+      "loss": 1.511,
+      "step": 20233
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.655308961868286,
+      "learning_rate": 4.465397832408142e-05,
+      "loss": 1.3626,
+      "step": 20234
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.3053030967712402,
+      "learning_rate": 4.4653713983610896e-05,
+      "loss": 1.3892,
+      "step": 20235
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.53096604347229,
+      "learning_rate": 4.465344964314037e-05,
+      "loss": 1.2896,
+      "step": 20236
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8323334455490112,
+      "learning_rate": 4.465318530266984e-05,
+      "loss": 1.3638,
+      "step": 20237
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9227359294891357,
+      "learning_rate": 4.465292096219931e-05,
+      "loss": 1.2844,
+      "step": 20238
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.804544448852539,
+      "learning_rate": 4.465265662172879e-05,
+      "loss": 1.397,
+      "step": 20239
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8986375331878662,
+      "learning_rate": 4.4652392281258264e-05,
+      "loss": 2.1014,
+      "step": 20240
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.750099778175354,
+      "learning_rate": 4.465212794078774e-05,
+      "loss": 1.1469,
+      "step": 20241
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7572872638702393,
+      "learning_rate": 4.465186360031721e-05,
+      "loss": 1.2463,
+      "step": 20242
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6725367307662964,
+      "learning_rate": 4.4651599259846684e-05,
+      "loss": 1.1843,
+      "step": 20243
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7906508445739746,
+      "learning_rate": 4.465133491937616e-05,
+      "loss": 1.2545,
+      "step": 20244
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0350940227508545,
+      "learning_rate": 4.465107057890564e-05,
+      "loss": 1.1935,
+      "step": 20245
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2541754245758057,
+      "learning_rate": 4.4650806238435105e-05,
+      "loss": 1.3843,
+      "step": 20246
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2129032611846924,
+      "learning_rate": 4.465054189796458e-05,
+      "loss": 2.3354,
+      "step": 20247
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0272791385650635,
+      "learning_rate": 4.465027755749405e-05,
+      "loss": 1.2415,
+      "step": 20248
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8501631021499634,
+      "learning_rate": 4.4650013217023526e-05,
+      "loss": 1.3624,
+      "step": 20249
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6832481622695923,
+      "learning_rate": 4.4649748876553006e-05,
+      "loss": 1.2162,
+      "step": 20250
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.613297939300537,
+      "learning_rate": 4.464948453608247e-05,
+      "loss": 1.2793,
+      "step": 20251
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8941818475723267,
+      "learning_rate": 4.464922019561195e-05,
+      "loss": 1.2312,
+      "step": 20252
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.287322759628296,
+      "learning_rate": 4.464895585514142e-05,
+      "loss": 1.465,
+      "step": 20253
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.352085828781128,
+      "learning_rate": 4.46486915146709e-05,
+      "loss": 1.4058,
+      "step": 20254
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9564275741577148,
+      "learning_rate": 4.4648427174200374e-05,
+      "loss": 1.3897,
+      "step": 20255
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7053941488265991,
+      "learning_rate": 4.464816283372985e-05,
+      "loss": 1.1605,
+      "step": 20256
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.637376308441162,
+      "learning_rate": 4.464789849325932e-05,
+      "loss": 1.566,
+      "step": 20257
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9754282236099243,
+      "learning_rate": 4.4647634152788795e-05,
+      "loss": 1.2307,
+      "step": 20258
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.4687016010284424,
+      "learning_rate": 4.464736981231827e-05,
+      "loss": 1.2538,
+      "step": 20259
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0481138229370117,
+      "learning_rate": 4.464710547184774e-05,
+      "loss": 1.4341,
+      "step": 20260
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9282034635543823,
+      "learning_rate": 4.4646841131377215e-05,
+      "loss": 1.1348,
+      "step": 20261
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.940371036529541,
+      "learning_rate": 4.464657679090669e-05,
+      "loss": 1.3239,
+      "step": 20262
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8885865211486816,
+      "learning_rate": 4.464631245043616e-05,
+      "loss": 1.2935,
+      "step": 20263
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.3565988540649414,
+      "learning_rate": 4.4646048109965636e-05,
+      "loss": 1.485,
+      "step": 20264
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8862260580062866,
+      "learning_rate": 4.4645783769495116e-05,
+      "loss": 1.1605,
+      "step": 20265
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9382587671279907,
+      "learning_rate": 4.464551942902458e-05,
+      "loss": 1.2858,
+      "step": 20266
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9830986261367798,
+      "learning_rate": 4.4645255088554064e-05,
+      "loss": 1.3535,
+      "step": 20267
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8098450899124146,
+      "learning_rate": 4.464499074808353e-05,
+      "loss": 1.322,
+      "step": 20268
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.156266450881958,
+      "learning_rate": 4.464472640761301e-05,
+      "loss": 1.0318,
+      "step": 20269
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0904812812805176,
+      "learning_rate": 4.4644462067142484e-05,
+      "loss": 1.263,
+      "step": 20270
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4323129653930664,
+      "learning_rate": 4.464419772667196e-05,
+      "loss": 1.2633,
+      "step": 20271
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.818034052848816,
+      "learning_rate": 4.464393338620143e-05,
+      "loss": 1.2929,
+      "step": 20272
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.5831661224365234,
+      "learning_rate": 4.46436690457309e-05,
+      "loss": 1.3237,
+      "step": 20273
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7661548852920532,
+      "learning_rate": 4.464340470526038e-05,
+      "loss": 1.4302,
+      "step": 20274
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7585197687149048,
+      "learning_rate": 4.464314036478985e-05,
+      "loss": 1.374,
+      "step": 20275
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8129210472106934,
+      "learning_rate": 4.4642876024319326e-05,
+      "loss": 1.3033,
+      "step": 20276
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1689813137054443,
+      "learning_rate": 4.46426116838488e-05,
+      "loss": 1.3606,
+      "step": 20277
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8199794292449951,
+      "learning_rate": 4.464234734337827e-05,
+      "loss": 1.2905,
+      "step": 20278
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9880319833755493,
+      "learning_rate": 4.4642083002907746e-05,
+      "loss": 1.3519,
+      "step": 20279
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0205459594726562,
+      "learning_rate": 4.464181866243723e-05,
+      "loss": 1.2496,
+      "step": 20280
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9267281293869019,
+      "learning_rate": 4.4641554321966694e-05,
+      "loss": 1.1639,
+      "step": 20281
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.9605581760406494,
+      "learning_rate": 4.464128998149617e-05,
+      "loss": 1.2834,
+      "step": 20282
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6292593479156494,
+      "learning_rate": 4.464102564102564e-05,
+      "loss": 1.9332,
+      "step": 20283
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.101022481918335,
+      "learning_rate": 4.4640761300555114e-05,
+      "loss": 1.3141,
+      "step": 20284
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7804744243621826,
+      "learning_rate": 4.4640496960084595e-05,
+      "loss": 1.2533,
+      "step": 20285
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9268819093704224,
+      "learning_rate": 4.464023261961406e-05,
+      "loss": 1.2793,
+      "step": 20286
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.62086820602417,
+      "learning_rate": 4.463996827914354e-05,
+      "loss": 1.4101,
+      "step": 20287
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7057409286499023,
+      "learning_rate": 4.463970393867301e-05,
+      "loss": 1.32,
+      "step": 20288
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5850226879119873,
+      "learning_rate": 4.463943959820249e-05,
+      "loss": 1.234,
+      "step": 20289
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8964673280715942,
+      "learning_rate": 4.463917525773196e-05,
+      "loss": 1.3065,
+      "step": 20290
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4163074493408203,
+      "learning_rate": 4.4638910917261436e-05,
+      "loss": 1.3128,
+      "step": 20291
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.765763521194458,
+      "learning_rate": 4.463864657679091e-05,
+      "loss": 1.2788,
+      "step": 20292
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8102657794952393,
+      "learning_rate": 4.463838223632038e-05,
+      "loss": 1.2004,
+      "step": 20293
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.375389814376831,
+      "learning_rate": 4.463811789584986e-05,
+      "loss": 1.2601,
+      "step": 20294
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0279738903045654,
+      "learning_rate": 4.463785355537933e-05,
+      "loss": 1.3471,
+      "step": 20295
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.9137449264526367,
+      "learning_rate": 4.4637589214908804e-05,
+      "loss": 1.3115,
+      "step": 20296
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.681252360343933,
+      "learning_rate": 4.463732487443828e-05,
+      "loss": 1.3041,
+      "step": 20297
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9910633563995361,
+      "learning_rate": 4.463706053396775e-05,
+      "loss": 1.3249,
+      "step": 20298
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.3971474170684814,
+      "learning_rate": 4.4636796193497225e-05,
+      "loss": 1.3273,
+      "step": 20299
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.177307605743408,
+      "learning_rate": 4.4636531853026705e-05,
+      "loss": 1.3362,
+      "step": 20300
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2845592498779297,
+      "learning_rate": 4.463626751255617e-05,
+      "loss": 1.4563,
+      "step": 20301
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1743595600128174,
+      "learning_rate": 4.463600317208565e-05,
+      "loss": 1.3291,
+      "step": 20302
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9000805616378784,
+      "learning_rate": 4.463573883161512e-05,
+      "loss": 1.2553,
+      "step": 20303
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.231431722640991,
+      "learning_rate": 4.46354744911446e-05,
+      "loss": 1.3885,
+      "step": 20304
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6452488899230957,
+      "learning_rate": 4.463521015067407e-05,
+      "loss": 1.1745,
+      "step": 20305
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.3270180225372314,
+      "learning_rate": 4.4634945810203546e-05,
+      "loss": 1.3078,
+      "step": 20306
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.175619602203369,
+      "learning_rate": 4.463468146973302e-05,
+      "loss": 1.4559,
+      "step": 20307
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.730905532836914,
+      "learning_rate": 4.4634417129262494e-05,
+      "loss": 1.2378,
+      "step": 20308
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.731414556503296,
+      "learning_rate": 4.463415278879197e-05,
+      "loss": 1.1863,
+      "step": 20309
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.5198140144348145,
+      "learning_rate": 4.463388844832144e-05,
+      "loss": 1.2738,
+      "step": 20310
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2436330318450928,
+      "learning_rate": 4.4633624107850914e-05,
+      "loss": 1.3209,
+      "step": 20311
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.298285484313965,
+      "learning_rate": 4.463335976738039e-05,
+      "loss": 1.3489,
+      "step": 20312
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9677209854125977,
+      "learning_rate": 4.463309542690986e-05,
+      "loss": 1.2374,
+      "step": 20313
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7772691249847412,
+      "learning_rate": 4.4632831086439335e-05,
+      "loss": 1.7247,
+      "step": 20314
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6955429315567017,
+      "learning_rate": 4.4632566745968815e-05,
+      "loss": 1.3306,
+      "step": 20315
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7270071506500244,
+      "learning_rate": 4.463230240549828e-05,
+      "loss": 1.1962,
+      "step": 20316
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.3485777378082275,
+      "learning_rate": 4.463203806502776e-05,
+      "loss": 1.246,
+      "step": 20317
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.970947504043579,
+      "learning_rate": 4.463177372455723e-05,
+      "loss": 1.3306,
+      "step": 20318
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.027531147003174,
+      "learning_rate": 4.46315093840867e-05,
+      "loss": 1.3168,
+      "step": 20319
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7420687675476074,
+      "learning_rate": 4.463124504361618e-05,
+      "loss": 1.1762,
+      "step": 20320
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.735252857208252,
+      "learning_rate": 4.463098070314565e-05,
+      "loss": 1.2866,
+      "step": 20321
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.931463599205017,
+      "learning_rate": 4.463071636267513e-05,
+      "loss": 1.111,
+      "step": 20322
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.1538567543029785,
+      "learning_rate": 4.46304520222046e-05,
+      "loss": 1.3406,
+      "step": 20323
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.74354887008667,
+      "learning_rate": 4.463018768173408e-05,
+      "loss": 1.2837,
+      "step": 20324
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0910401344299316,
+      "learning_rate": 4.462992334126355e-05,
+      "loss": 1.2277,
+      "step": 20325
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0710837841033936,
+      "learning_rate": 4.4629659000793025e-05,
+      "loss": 1.4998,
+      "step": 20326
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.001377820968628,
+      "learning_rate": 4.46293946603225e-05,
+      "loss": 1.2566,
+      "step": 20327
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.520345687866211,
+      "learning_rate": 4.462913031985197e-05,
+      "loss": 1.3446,
+      "step": 20328
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9950098991394043,
+      "learning_rate": 4.4628865979381445e-05,
+      "loss": 1.3181,
+      "step": 20329
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8173792362213135,
+      "learning_rate": 4.462860163891092e-05,
+      "loss": 1.3313,
+      "step": 20330
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.084810972213745,
+      "learning_rate": 4.462833729844039e-05,
+      "loss": 1.3258,
+      "step": 20331
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7633014917373657,
+      "learning_rate": 4.4628072957969866e-05,
+      "loss": 1.0837,
+      "step": 20332
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8451135158538818,
+      "learning_rate": 4.462780861749934e-05,
+      "loss": 1.3499,
+      "step": 20333
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.743551254272461,
+      "learning_rate": 4.462754427702881e-05,
+      "loss": 1.1863,
+      "step": 20334
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.972224235534668,
+      "learning_rate": 4.4627279936558293e-05,
+      "loss": 1.0796,
+      "step": 20335
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.10683274269104,
+      "learning_rate": 4.462701559608776e-05,
+      "loss": 1.1369,
+      "step": 20336
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7122172117233276,
+      "learning_rate": 4.462675125561724e-05,
+      "loss": 1.2054,
+      "step": 20337
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.775611400604248,
+      "learning_rate": 4.462648691514671e-05,
+      "loss": 1.16,
+      "step": 20338
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3977773189544678,
+      "learning_rate": 4.462622257467619e-05,
+      "loss": 1.3256,
+      "step": 20339
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.784874677658081,
+      "learning_rate": 4.462595823420566e-05,
+      "loss": 1.3645,
+      "step": 20340
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8618152141571045,
+      "learning_rate": 4.4625693893735135e-05,
+      "loss": 1.3339,
+      "step": 20341
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.56599497795105,
+      "learning_rate": 4.462542955326461e-05,
+      "loss": 1.2971,
+      "step": 20342
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8621317148208618,
+      "learning_rate": 4.462516521279408e-05,
+      "loss": 1.3803,
+      "step": 20343
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.115192174911499,
+      "learning_rate": 4.4624900872323556e-05,
+      "loss": 1.3326,
+      "step": 20344
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.982906460762024,
+      "learning_rate": 4.462463653185303e-05,
+      "loss": 1.2653,
+      "step": 20345
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0570478439331055,
+      "learning_rate": 4.46243721913825e-05,
+      "loss": 1.3568,
+      "step": 20346
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.580852746963501,
+      "learning_rate": 4.4624107850911976e-05,
+      "loss": 1.2623,
+      "step": 20347
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7732473611831665,
+      "learning_rate": 4.462384351044145e-05,
+      "loss": 1.3232,
+      "step": 20348
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.023280620574951,
+      "learning_rate": 4.4623579169970923e-05,
+      "loss": 1.4105,
+      "step": 20349
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.006627082824707,
+      "learning_rate": 4.4623314829500404e-05,
+      "loss": 1.3881,
+      "step": 20350
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.174321413040161,
+      "learning_rate": 4.462305048902987e-05,
+      "loss": 1.1267,
+      "step": 20351
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.045100688934326,
+      "learning_rate": 4.462278614855935e-05,
+      "loss": 1.1846,
+      "step": 20352
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.340850353240967,
+      "learning_rate": 4.462252180808882e-05,
+      "loss": 1.2274,
+      "step": 20353
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9639098644256592,
+      "learning_rate": 4.46222574676183e-05,
+      "loss": 1.2527,
+      "step": 20354
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.108236789703369,
+      "learning_rate": 4.462199312714777e-05,
+      "loss": 1.3588,
+      "step": 20355
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9588137865066528,
+      "learning_rate": 4.462172878667724e-05,
+      "loss": 1.2094,
+      "step": 20356
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0848002433776855,
+      "learning_rate": 4.462146444620672e-05,
+      "loss": 1.3361,
+      "step": 20357
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3106791973114014,
+      "learning_rate": 4.4621200105736186e-05,
+      "loss": 1.3012,
+      "step": 20358
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9334908723831177,
+      "learning_rate": 4.4620935765265666e-05,
+      "loss": 1.32,
+      "step": 20359
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.129425287246704,
+      "learning_rate": 4.462067142479514e-05,
+      "loss": 1.2424,
+      "step": 20360
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3550806045532227,
+      "learning_rate": 4.462040708432461e-05,
+      "loss": 1.7119,
+      "step": 20361
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3270926475524902,
+      "learning_rate": 4.462014274385409e-05,
+      "loss": 1.4137,
+      "step": 20362
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.363349676132202,
+      "learning_rate": 4.461987840338356e-05,
+      "loss": 1.4448,
+      "step": 20363
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8695006370544434,
+      "learning_rate": 4.4619614062913034e-05,
+      "loss": 1.4421,
+      "step": 20364
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6874741315841675,
+      "learning_rate": 4.461934972244251e-05,
+      "loss": 1.2469,
+      "step": 20365
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9404882192611694,
+      "learning_rate": 4.461908538197198e-05,
+      "loss": 1.3292,
+      "step": 20366
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2742855548858643,
+      "learning_rate": 4.4618821041501455e-05,
+      "loss": 1.3545,
+      "step": 20367
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8068606853485107,
+      "learning_rate": 4.461855670103093e-05,
+      "loss": 2.0783,
+      "step": 20368
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0746665000915527,
+      "learning_rate": 4.46182923605604e-05,
+      "loss": 1.4045,
+      "step": 20369
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.397578716278076,
+      "learning_rate": 4.461802802008988e-05,
+      "loss": 1.3404,
+      "step": 20370
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3648269176483154,
+      "learning_rate": 4.461776367961935e-05,
+      "loss": 1.2425,
+      "step": 20371
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9877781867980957,
+      "learning_rate": 4.461749933914883e-05,
+      "loss": 1.3556,
+      "step": 20372
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1117587089538574,
+      "learning_rate": 4.4617234998678296e-05,
+      "loss": 1.232,
+      "step": 20373
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7725459337234497,
+      "learning_rate": 4.4616970658207776e-05,
+      "loss": 1.8912,
+      "step": 20374
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8027647733688354,
+      "learning_rate": 4.461670631773725e-05,
+      "loss": 1.1835,
+      "step": 20375
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.5715456008911133,
+      "learning_rate": 4.4616441977266723e-05,
+      "loss": 1.501,
+      "step": 20376
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.5886473655700684,
+      "learning_rate": 4.46161776367962e-05,
+      "loss": 1.4944,
+      "step": 20377
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0017852783203125,
+      "learning_rate": 4.461591329632567e-05,
+      "loss": 1.2196,
+      "step": 20378
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8167729377746582,
+      "learning_rate": 4.4615648955855144e-05,
+      "loss": 1.2191,
+      "step": 20379
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0423967838287354,
+      "learning_rate": 4.461538461538462e-05,
+      "loss": 1.21,
+      "step": 20380
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6497832536697388,
+      "learning_rate": 4.461512027491409e-05,
+      "loss": 1.2917,
+      "step": 20381
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.283242702484131,
+      "learning_rate": 4.4614855934443565e-05,
+      "loss": 1.2697,
+      "step": 20382
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9386069774627686,
+      "learning_rate": 4.461459159397304e-05,
+      "loss": 1.3981,
+      "step": 20383
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6703672409057617,
+      "learning_rate": 4.461432725350251e-05,
+      "loss": 1.2879,
+      "step": 20384
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8186514377593994,
+      "learning_rate": 4.461406291303199e-05,
+      "loss": 1.1536,
+      "step": 20385
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9033864736557007,
+      "learning_rate": 4.461379857256146e-05,
+      "loss": 1.2921,
+      "step": 20386
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.910375952720642,
+      "learning_rate": 4.461353423209094e-05,
+      "loss": 1.3019,
+      "step": 20387
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1408777236938477,
+      "learning_rate": 4.4613269891620406e-05,
+      "loss": 1.3736,
+      "step": 20388
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8562883138656616,
+      "learning_rate": 4.461300555114989e-05,
+      "loss": 1.248,
+      "step": 20389
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1788222789764404,
+      "learning_rate": 4.461274121067936e-05,
+      "loss": 1.3959,
+      "step": 20390
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8958722352981567,
+      "learning_rate": 4.461247687020883e-05,
+      "loss": 1.1949,
+      "step": 20391
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.080841541290283,
+      "learning_rate": 4.461221252973831e-05,
+      "loss": 1.3771,
+      "step": 20392
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9767078161239624,
+      "learning_rate": 4.4611948189267774e-05,
+      "loss": 1.5039,
+      "step": 20393
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.915259838104248,
+      "learning_rate": 4.4611683848797255e-05,
+      "loss": 1.2455,
+      "step": 20394
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6496151685714722,
+      "learning_rate": 4.461141950832673e-05,
+      "loss": 1.3465,
+      "step": 20395
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.6273117065429688,
+      "learning_rate": 4.46111551678562e-05,
+      "loss": 1.2574,
+      "step": 20396
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0593268871307373,
+      "learning_rate": 4.4610890827385675e-05,
+      "loss": 1.3131,
+      "step": 20397
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.3238530158996582,
+      "learning_rate": 4.461062648691515e-05,
+      "loss": 1.587,
+      "step": 20398
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0041329860687256,
+      "learning_rate": 4.461036214644462e-05,
+      "loss": 1.2657,
+      "step": 20399
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.5247156620025635,
+      "learning_rate": 4.4610097805974096e-05,
+      "loss": 1.206,
+      "step": 20400
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6441221237182617,
+      "learning_rate": 4.460983346550357e-05,
+      "loss": 1.4007,
+      "step": 20401
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1566944122314453,
+      "learning_rate": 4.460956912503304e-05,
+      "loss": 1.1864,
+      "step": 20402
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2624313831329346,
+      "learning_rate": 4.460930478456252e-05,
+      "loss": 1.3239,
+      "step": 20403
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1352367401123047,
+      "learning_rate": 4.460904044409199e-05,
+      "loss": 1.3683,
+      "step": 20404
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1316752433776855,
+      "learning_rate": 4.460877610362147e-05,
+      "loss": 1.4064,
+      "step": 20405
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.4680519104003906,
+      "learning_rate": 4.460851176315094e-05,
+      "loss": 1.2819,
+      "step": 20406
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.08895206451416,
+      "learning_rate": 4.460824742268042e-05,
+      "loss": 1.298,
+      "step": 20407
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6514078378677368,
+      "learning_rate": 4.4607983082209884e-05,
+      "loss": 1.2551,
+      "step": 20408
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.7440454959869385,
+      "learning_rate": 4.4607718741739365e-05,
+      "loss": 1.2978,
+      "step": 20409
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.122202157974243,
+      "learning_rate": 4.460745440126884e-05,
+      "loss": 1.3423,
+      "step": 20410
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.4801125526428223,
+      "learning_rate": 4.460719006079831e-05,
+      "loss": 1.4042,
+      "step": 20411
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2197375297546387,
+      "learning_rate": 4.4606925720327786e-05,
+      "loss": 1.3343,
+      "step": 20412
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9182935953140259,
+      "learning_rate": 4.460666137985726e-05,
+      "loss": 1.3304,
+      "step": 20413
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.467627763748169,
+      "learning_rate": 4.460639703938673e-05,
+      "loss": 1.2662,
+      "step": 20414
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9995297193527222,
+      "learning_rate": 4.4606132698916206e-05,
+      "loss": 1.4247,
+      "step": 20415
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9166357517242432,
+      "learning_rate": 4.460586835844568e-05,
+      "loss": 1.3069,
+      "step": 20416
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9476526975631714,
+      "learning_rate": 4.4605604017975153e-05,
+      "loss": 1.3273,
+      "step": 20417
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6821950674057007,
+      "learning_rate": 4.460533967750463e-05,
+      "loss": 1.2459,
+      "step": 20418
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6925170421600342,
+      "learning_rate": 4.46050753370341e-05,
+      "loss": 1.1614,
+      "step": 20419
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.050232410430908,
+      "learning_rate": 4.460481099656358e-05,
+      "loss": 1.3413,
+      "step": 20420
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7611327171325684,
+      "learning_rate": 4.460454665609305e-05,
+      "loss": 1.1371,
+      "step": 20421
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9080809354782104,
+      "learning_rate": 4.460428231562253e-05,
+      "loss": 1.2473,
+      "step": 20422
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2700095176696777,
+      "learning_rate": 4.4604017975151995e-05,
+      "loss": 1.3179,
+      "step": 20423
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.258225440979004,
+      "learning_rate": 4.4603753634681475e-05,
+      "loss": 1.4089,
+      "step": 20424
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.051508903503418,
+      "learning_rate": 4.460348929421095e-05,
+      "loss": 1.3546,
+      "step": 20425
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0251238346099854,
+      "learning_rate": 4.460322495374042e-05,
+      "loss": 1.2327,
+      "step": 20426
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.115898370742798,
+      "learning_rate": 4.4602960613269896e-05,
+      "loss": 1.3755,
+      "step": 20427
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.7275407314300537,
+      "learning_rate": 4.460269627279936e-05,
+      "loss": 1.279,
+      "step": 20428
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2151429653167725,
+      "learning_rate": 4.460243193232884e-05,
+      "loss": 1.3367,
+      "step": 20429
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.864126443862915,
+      "learning_rate": 4.4602167591858317e-05,
+      "loss": 1.2492,
+      "step": 20430
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.6263458728790283,
+      "learning_rate": 4.460190325138779e-05,
+      "loss": 1.4152,
+      "step": 20431
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8077582120895386,
+      "learning_rate": 4.4601638910917264e-05,
+      "loss": 1.4046,
+      "step": 20432
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9503967761993408,
+      "learning_rate": 4.460137457044674e-05,
+      "loss": 1.2893,
+      "step": 20433
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.47934627532959,
+      "learning_rate": 4.460111022997621e-05,
+      "loss": 1.5482,
+      "step": 20434
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2728304862976074,
+      "learning_rate": 4.460084588950569e-05,
+      "loss": 1.4578,
+      "step": 20435
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0732271671295166,
+      "learning_rate": 4.460058154903516e-05,
+      "loss": 1.441,
+      "step": 20436
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.18400502204895,
+      "learning_rate": 4.460031720856463e-05,
+      "loss": 1.2634,
+      "step": 20437
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3638784885406494,
+      "learning_rate": 4.4600052868094105e-05,
+      "loss": 1.4905,
+      "step": 20438
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8055957555770874,
+      "learning_rate": 4.459978852762358e-05,
+      "loss": 1.2305,
+      "step": 20439
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.200408935546875,
+      "learning_rate": 4.459952418715306e-05,
+      "loss": 1.3925,
+      "step": 20440
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8035131692886353,
+      "learning_rate": 4.4599259846682526e-05,
+      "loss": 1.1941,
+      "step": 20441
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2598206996917725,
+      "learning_rate": 4.4598995506212006e-05,
+      "loss": 1.4471,
+      "step": 20442
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.0954837799072266,
+      "learning_rate": 4.459873116574147e-05,
+      "loss": 1.4926,
+      "step": 20443
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.5413951873779297,
+      "learning_rate": 4.459846682527095e-05,
+      "loss": 1.2995,
+      "step": 20444
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.880098581314087,
+      "learning_rate": 4.459820248480043e-05,
+      "loss": 1.5672,
+      "step": 20445
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.342653274536133,
+      "learning_rate": 4.45979381443299e-05,
+      "loss": 1.38,
+      "step": 20446
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.4275476932525635,
+      "learning_rate": 4.4597673803859374e-05,
+      "loss": 1.2615,
+      "step": 20447
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.5116188526153564,
+      "learning_rate": 4.459740946338885e-05,
+      "loss": 1.7661,
+      "step": 20448
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3628687858581543,
+      "learning_rate": 4.459714512291832e-05,
+      "loss": 1.3739,
+      "step": 20449
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9771831035614014,
+      "learning_rate": 4.4596880782447795e-05,
+      "loss": 1.3848,
+      "step": 20450
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.373323440551758,
+      "learning_rate": 4.459661644197727e-05,
+      "loss": 1.4673,
+      "step": 20451
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3981566429138184,
+      "learning_rate": 4.459635210150674e-05,
+      "loss": 1.3403,
+      "step": 20452
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6874005794525146,
+      "learning_rate": 4.4596087761036216e-05,
+      "loss": 1.2968,
+      "step": 20453
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7147164344787598,
+      "learning_rate": 4.459582342056569e-05,
+      "loss": 1.1878,
+      "step": 20454
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.5949376821517944,
+      "learning_rate": 4.459555908009517e-05,
+      "loss": 1.1629,
+      "step": 20455
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8874484300613403,
+      "learning_rate": 4.4595294739624636e-05,
+      "loss": 1.3069,
+      "step": 20456
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.8707144260406494,
+      "learning_rate": 4.4595030399154117e-05,
+      "loss": 1.3611,
+      "step": 20457
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3108694553375244,
+      "learning_rate": 4.459476605868358e-05,
+      "loss": 1.2445,
+      "step": 20458
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.4040322303771973,
+      "learning_rate": 4.4594501718213064e-05,
+      "loss": 1.3757,
+      "step": 20459
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.736203670501709,
+      "learning_rate": 4.459423737774254e-05,
+      "loss": 1.5013,
+      "step": 20460
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.164214611053467,
+      "learning_rate": 4.459397303727201e-05,
+      "loss": 1.3489,
+      "step": 20461
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9973722696304321,
+      "learning_rate": 4.4593708696801484e-05,
+      "loss": 1.4303,
+      "step": 20462
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9207035303115845,
+      "learning_rate": 4.459344435633095e-05,
+      "loss": 1.278,
+      "step": 20463
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9507875442504883,
+      "learning_rate": 4.459318001586043e-05,
+      "loss": 1.3108,
+      "step": 20464
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8637105226516724,
+      "learning_rate": 4.4592915675389905e-05,
+      "loss": 1.2858,
+      "step": 20465
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6774039268493652,
+      "learning_rate": 4.459265133491938e-05,
+      "loss": 2.0108,
+      "step": 20466
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6497507095336914,
+      "learning_rate": 4.459238699444885e-05,
+      "loss": 1.1501,
+      "step": 20467
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.814946174621582,
+      "learning_rate": 4.4592122653978326e-05,
+      "loss": 1.3405,
+      "step": 20468
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9023431539535522,
+      "learning_rate": 4.45918583135078e-05,
+      "loss": 1.3724,
+      "step": 20469
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9889503717422485,
+      "learning_rate": 4.459159397303728e-05,
+      "loss": 1.2593,
+      "step": 20470
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7464615106582642,
+      "learning_rate": 4.4591329632566747e-05,
+      "loss": 1.248,
+      "step": 20471
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.965201497077942,
+      "learning_rate": 4.459106529209622e-05,
+      "loss": 1.3619,
+      "step": 20472
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.190993309020996,
+      "learning_rate": 4.4590800951625694e-05,
+      "loss": 1.2707,
+      "step": 20473
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9238684177398682,
+      "learning_rate": 4.459053661115517e-05,
+      "loss": 1.4027,
+      "step": 20474
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0516321659088135,
+      "learning_rate": 4.459027227068465e-05,
+      "loss": 1.2059,
+      "step": 20475
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9967387914657593,
+      "learning_rate": 4.4590007930214114e-05,
+      "loss": 1.428,
+      "step": 20476
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.345332384109497,
+      "learning_rate": 4.4589743589743595e-05,
+      "loss": 1.4205,
+      "step": 20477
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.5260834693908691,
+      "learning_rate": 4.458947924927306e-05,
+      "loss": 1.128,
+      "step": 20478
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.445805072784424,
+      "learning_rate": 4.458921490880254e-05,
+      "loss": 1.3407,
+      "step": 20479
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.4589993953704834,
+      "learning_rate": 4.4588950568332015e-05,
+      "loss": 1.3108,
+      "step": 20480
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9499863386154175,
+      "learning_rate": 4.458868622786149e-05,
+      "loss": 1.4225,
+      "step": 20481
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7873153686523438,
+      "learning_rate": 4.458842188739096e-05,
+      "loss": 1.2318,
+      "step": 20482
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.803086757659912,
+      "learning_rate": 4.4588157546920436e-05,
+      "loss": 1.2634,
+      "step": 20483
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.4408369064331055,
+      "learning_rate": 4.458789320644991e-05,
+      "loss": 1.2659,
+      "step": 20484
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1849470138549805,
+      "learning_rate": 4.458762886597938e-05,
+      "loss": 1.2842,
+      "step": 20485
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.7451963424682617,
+      "learning_rate": 4.458736452550886e-05,
+      "loss": 1.3977,
+      "step": 20486
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1999599933624268,
+      "learning_rate": 4.458710018503833e-05,
+      "loss": 1.3521,
+      "step": 20487
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.4669349193572998,
+      "learning_rate": 4.4586835844567804e-05,
+      "loss": 1.8584,
+      "step": 20488
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.5592422485351562,
+      "learning_rate": 4.458657150409728e-05,
+      "loss": 1.2067,
+      "step": 20489
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.4301304817199707,
+      "learning_rate": 4.458630716362676e-05,
+      "loss": 1.529,
+      "step": 20490
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9454171657562256,
+      "learning_rate": 4.4586042823156225e-05,
+      "loss": 1.2014,
+      "step": 20491
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1688783168792725,
+      "learning_rate": 4.4585778482685705e-05,
+      "loss": 1.2687,
+      "step": 20492
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2076663970947266,
+      "learning_rate": 4.458551414221517e-05,
+      "loss": 1.2932,
+      "step": 20493
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7839014530181885,
+      "learning_rate": 4.458524980174465e-05,
+      "loss": 1.2583,
+      "step": 20494
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.766484260559082,
+      "learning_rate": 4.4584985461274126e-05,
+      "loss": 1.5563,
+      "step": 20495
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9051185846328735,
+      "learning_rate": 4.45847211208036e-05,
+      "loss": 1.3587,
+      "step": 20496
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7985368967056274,
+      "learning_rate": 4.458445678033307e-05,
+      "loss": 1.2611,
+      "step": 20497
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0260825157165527,
+      "learning_rate": 4.4584192439862547e-05,
+      "loss": 1.3508,
+      "step": 20498
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9843252897262573,
+      "learning_rate": 4.458392809939202e-05,
+      "loss": 1.3434,
+      "step": 20499
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.535771608352661,
+      "learning_rate": 4.4583663758921494e-05,
+      "loss": 1.5253,
+      "step": 20500
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6148396730422974,
+      "learning_rate": 4.458339941845097e-05,
+      "loss": 1.2403,
+      "step": 20501
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.5793659687042236,
+      "learning_rate": 4.458313507798044e-05,
+      "loss": 1.2658,
+      "step": 20502
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7944551706314087,
+      "learning_rate": 4.4582870737509914e-05,
+      "loss": 1.1717,
+      "step": 20503
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8191536664962769,
+      "learning_rate": 4.458260639703939e-05,
+      "loss": 1.2386,
+      "step": 20504
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0163049697875977,
+      "learning_rate": 4.458234205656887e-05,
+      "loss": 1.332,
+      "step": 20505
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.233095407485962,
+      "learning_rate": 4.4582077716098335e-05,
+      "loss": 1.3847,
+      "step": 20506
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.007297992706299,
+      "learning_rate": 4.4581813375627815e-05,
+      "loss": 1.3084,
+      "step": 20507
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0665550231933594,
+      "learning_rate": 4.458154903515728e-05,
+      "loss": 1.3514,
+      "step": 20508
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9407545328140259,
+      "learning_rate": 4.4581284694686756e-05,
+      "loss": 1.3998,
+      "step": 20509
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7771971225738525,
+      "learning_rate": 4.4581020354216236e-05,
+      "loss": 1.2938,
+      "step": 20510
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7797569036483765,
+      "learning_rate": 4.45807560137457e-05,
+      "loss": 1.3616,
+      "step": 20511
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8586312532424927,
+      "learning_rate": 4.458049167327518e-05,
+      "loss": 1.1946,
+      "step": 20512
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.848065972328186,
+      "learning_rate": 4.458022733280465e-05,
+      "loss": 1.2448,
+      "step": 20513
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9788286685943604,
+      "learning_rate": 4.457996299233413e-05,
+      "loss": 1.3095,
+      "step": 20514
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.713773012161255,
+      "learning_rate": 4.4579698651863604e-05,
+      "loss": 1.4162,
+      "step": 20515
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9149638414382935,
+      "learning_rate": 4.457943431139308e-05,
+      "loss": 2.0845,
+      "step": 20516
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.126089334487915,
+      "learning_rate": 4.457916997092255e-05,
+      "loss": 1.3183,
+      "step": 20517
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9782687425613403,
+      "learning_rate": 4.4578905630452025e-05,
+      "loss": 1.4225,
+      "step": 20518
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7421927452087402,
+      "learning_rate": 4.45786412899815e-05,
+      "loss": 1.2134,
+      "step": 20519
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.0130748748779297,
+      "learning_rate": 4.457837694951097e-05,
+      "loss": 1.4261,
+      "step": 20520
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.120161294937134,
+      "learning_rate": 4.4578112609040445e-05,
+      "loss": 1.9486,
+      "step": 20521
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.146470069885254,
+      "learning_rate": 4.457784826856992e-05,
+      "loss": 1.3631,
+      "step": 20522
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1576290130615234,
+      "learning_rate": 4.457758392809939e-05,
+      "loss": 1.2917,
+      "step": 20523
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7816699743270874,
+      "learning_rate": 4.4577319587628866e-05,
+      "loss": 1.2661,
+      "step": 20524
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1481316089630127,
+      "learning_rate": 4.4577055247158347e-05,
+      "loss": 1.2556,
+      "step": 20525
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6018550395965576,
+      "learning_rate": 4.457679090668781e-05,
+      "loss": 1.1952,
+      "step": 20526
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.831378698348999,
+      "learning_rate": 4.4576526566217294e-05,
+      "loss": 1.2221,
+      "step": 20527
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9136024713516235,
+      "learning_rate": 4.457626222574676e-05,
+      "loss": 2.1868,
+      "step": 20528
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.074307918548584,
+      "learning_rate": 4.457599788527624e-05,
+      "loss": 1.2197,
+      "step": 20529
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6547876596450806,
+      "learning_rate": 4.4575733544805714e-05,
+      "loss": 1.2423,
+      "step": 20530
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.4198079109191895,
+      "learning_rate": 4.457546920433519e-05,
+      "loss": 1.3003,
+      "step": 20531
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.786365270614624,
+      "learning_rate": 4.457520486386466e-05,
+      "loss": 1.2939,
+      "step": 20532
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8107593059539795,
+      "learning_rate": 4.4574940523394135e-05,
+      "loss": 1.2911,
+      "step": 20533
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.065835475921631,
+      "learning_rate": 4.457467618292361e-05,
+      "loss": 1.3278,
+      "step": 20534
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.6135365962982178,
+      "learning_rate": 4.457441184245308e-05,
+      "loss": 2.1629,
+      "step": 20535
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1809964179992676,
+      "learning_rate": 4.4574147501982556e-05,
+      "loss": 1.3634,
+      "step": 20536
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9673900604248047,
+      "learning_rate": 4.457388316151203e-05,
+      "loss": 1.16,
+      "step": 20537
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.574666738510132,
+      "learning_rate": 4.45736188210415e-05,
+      "loss": 1.3599,
+      "step": 20538
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9640921354293823,
+      "learning_rate": 4.4573354480570976e-05,
+      "loss": 1.1532,
+      "step": 20539
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7591429948806763,
+      "learning_rate": 4.457309014010046e-05,
+      "loss": 2.0253,
+      "step": 20540
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.9692625999450684,
+      "learning_rate": 4.4572825799629924e-05,
+      "loss": 1.2487,
+      "step": 20541
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8066679239273071,
+      "learning_rate": 4.4572561459159404e-05,
+      "loss": 1.2058,
+      "step": 20542
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.400956869125366,
+      "learning_rate": 4.457229711868887e-05,
+      "loss": 1.4399,
+      "step": 20543
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.152830123901367,
+      "learning_rate": 4.457203277821835e-05,
+      "loss": 1.3442,
+      "step": 20544
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0153026580810547,
+      "learning_rate": 4.4571768437747825e-05,
+      "loss": 1.3897,
+      "step": 20545
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.818893313407898,
+      "learning_rate": 4.457150409727729e-05,
+      "loss": 1.2497,
+      "step": 20546
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7033227682113647,
+      "learning_rate": 4.457123975680677e-05,
+      "loss": 1.2728,
+      "step": 20547
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.8270516395568848,
+      "learning_rate": 4.457097541633624e-05,
+      "loss": 1.4625,
+      "step": 20548
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7614963054656982,
+      "learning_rate": 4.457071107586572e-05,
+      "loss": 1.3006,
+      "step": 20549
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.0733418464660645,
+      "learning_rate": 4.457044673539519e-05,
+      "loss": 1.5339,
+      "step": 20550
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0644900798797607,
+      "learning_rate": 4.4570182394924666e-05,
+      "loss": 1.4259,
+      "step": 20551
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.669647693634033,
+      "learning_rate": 4.456991805445414e-05,
+      "loss": 1.3868,
+      "step": 20552
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1931276321411133,
+      "learning_rate": 4.456965371398361e-05,
+      "loss": 1.2979,
+      "step": 20553
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.548604726791382,
+      "learning_rate": 4.456938937351309e-05,
+      "loss": 1.2525,
+      "step": 20554
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.8540561199188232,
+      "learning_rate": 4.456912503304256e-05,
+      "loss": 1.4129,
+      "step": 20555
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.032524824142456,
+      "learning_rate": 4.4568860692572034e-05,
+      "loss": 1.304,
+      "step": 20556
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0161328315734863,
+      "learning_rate": 4.456859635210151e-05,
+      "loss": 1.3168,
+      "step": 20557
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6246808767318726,
+      "learning_rate": 4.456833201163098e-05,
+      "loss": 1.8521,
+      "step": 20558
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9568569660186768,
+      "learning_rate": 4.4568067671160455e-05,
+      "loss": 1.2673,
+      "step": 20559
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.141505002975464,
+      "learning_rate": 4.4567803330689935e-05,
+      "loss": 1.4426,
+      "step": 20560
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.0332789421081543,
+      "learning_rate": 4.45675389902194e-05,
+      "loss": 1.5594,
+      "step": 20561
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.903911828994751,
+      "learning_rate": 4.456727464974888e-05,
+      "loss": 1.4303,
+      "step": 20562
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7743149995803833,
+      "learning_rate": 4.456701030927835e-05,
+      "loss": 1.1956,
+      "step": 20563
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.5435869693756104,
+      "learning_rate": 4.456674596880783e-05,
+      "loss": 1.8107,
+      "step": 20564
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.079515218734741,
+      "learning_rate": 4.45664816283373e-05,
+      "loss": 1.2839,
+      "step": 20565
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.350980520248413,
+      "learning_rate": 4.4566217287866776e-05,
+      "loss": 1.5008,
+      "step": 20566
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.829593539237976,
+      "learning_rate": 4.456595294739625e-05,
+      "loss": 1.1595,
+      "step": 20567
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8207811117172241,
+      "learning_rate": 4.4565688606925724e-05,
+      "loss": 1.2596,
+      "step": 20568
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.7226717472076416,
+      "learning_rate": 4.45654242664552e-05,
+      "loss": 1.3597,
+      "step": 20569
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.181306838989258,
+      "learning_rate": 4.456515992598467e-05,
+      "loss": 1.4174,
+      "step": 20570
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8951144218444824,
+      "learning_rate": 4.4564895585514144e-05,
+      "loss": 1.2423,
+      "step": 20571
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9658300876617432,
+      "learning_rate": 4.456463124504362e-05,
+      "loss": 1.27,
+      "step": 20572
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.683528184890747,
+      "learning_rate": 4.456436690457309e-05,
+      "loss": 1.2706,
+      "step": 20573
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7493184804916382,
+      "learning_rate": 4.4564102564102565e-05,
+      "loss": 1.3037,
+      "step": 20574
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8530288934707642,
+      "learning_rate": 4.4563838223632045e-05,
+      "loss": 1.3194,
+      "step": 20575
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.8659188747406006,
+      "learning_rate": 4.456357388316151e-05,
+      "loss": 1.4839,
+      "step": 20576
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.307107448577881,
+      "learning_rate": 4.456330954269099e-05,
+      "loss": 1.3589,
+      "step": 20577
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8358566761016846,
+      "learning_rate": 4.456304520222046e-05,
+      "loss": 1.2668,
+      "step": 20578
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.4518561363220215,
+      "learning_rate": 4.456278086174994e-05,
+      "loss": 1.0994,
+      "step": 20579
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.2348055839538574,
+      "learning_rate": 4.456251652127941e-05,
+      "loss": 1.3569,
+      "step": 20580
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8085672855377197,
+      "learning_rate": 4.456225218080888e-05,
+      "loss": 1.2252,
+      "step": 20581
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.717812418937683,
+      "learning_rate": 4.456198784033836e-05,
+      "loss": 1.1596,
+      "step": 20582
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.766401767730713,
+      "learning_rate": 4.456172349986783e-05,
+      "loss": 1.2575,
+      "step": 20583
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.413280487060547,
+      "learning_rate": 4.456145915939731e-05,
+      "loss": 1.4461,
+      "step": 20584
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4411444664001465,
+      "learning_rate": 4.456119481892678e-05,
+      "loss": 1.2322,
+      "step": 20585
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0828545093536377,
+      "learning_rate": 4.4560930478456255e-05,
+      "loss": 1.3485,
+      "step": 20586
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.821913719177246,
+      "learning_rate": 4.456066613798573e-05,
+      "loss": 1.3929,
+      "step": 20587
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.356860876083374,
+      "learning_rate": 4.45604017975152e-05,
+      "loss": 1.4024,
+      "step": 20588
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8697904348373413,
+      "learning_rate": 4.4560137457044675e-05,
+      "loss": 1.2117,
+      "step": 20589
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8842060565948486,
+      "learning_rate": 4.455987311657415e-05,
+      "loss": 1.2776,
+      "step": 20590
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.127439022064209,
+      "learning_rate": 4.455960877610362e-05,
+      "loss": 1.1986,
+      "step": 20591
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9792416095733643,
+      "learning_rate": 4.4559344435633096e-05,
+      "loss": 1.4026,
+      "step": 20592
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7481412887573242,
+      "learning_rate": 4.455908009516257e-05,
+      "loss": 1.3418,
+      "step": 20593
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.2397549152374268,
+      "learning_rate": 4.455881575469204e-05,
+      "loss": 1.3931,
+      "step": 20594
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8727649450302124,
+      "learning_rate": 4.4558551414221524e-05,
+      "loss": 1.4268,
+      "step": 20595
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.2308454513549805,
+      "learning_rate": 4.455828707375099e-05,
+      "loss": 1.3615,
+      "step": 20596
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.702401638031006,
+      "learning_rate": 4.455802273328047e-05,
+      "loss": 1.284,
+      "step": 20597
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.075160503387451,
+      "learning_rate": 4.455775839280994e-05,
+      "loss": 1.3695,
+      "step": 20598
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4221372604370117,
+      "learning_rate": 4.455749405233942e-05,
+      "loss": 1.3814,
+      "step": 20599
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.241407632827759,
+      "learning_rate": 4.455722971186889e-05,
+      "loss": 1.2555,
+      "step": 20600
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0653934478759766,
+      "learning_rate": 4.4556965371398365e-05,
+      "loss": 1.2926,
+      "step": 20601
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7330889701843262,
+      "learning_rate": 4.455670103092784e-05,
+      "loss": 1.1433,
+      "step": 20602
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8215413093566895,
+      "learning_rate": 4.455643669045731e-05,
+      "loss": 1.6506,
+      "step": 20603
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8890647888183594,
+      "learning_rate": 4.4556172349986786e-05,
+      "loss": 1.3791,
+      "step": 20604
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.23030948638916,
+      "learning_rate": 4.455590800951626e-05,
+      "loss": 1.3883,
+      "step": 20605
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1490843296051025,
+      "learning_rate": 4.455564366904573e-05,
+      "loss": 1.4004,
+      "step": 20606
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.888370394706726,
+      "learning_rate": 4.4555379328575206e-05,
+      "loss": 1.2299,
+      "step": 20607
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9204579591751099,
+      "learning_rate": 4.455511498810468e-05,
+      "loss": 1.305,
+      "step": 20608
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.2005293369293213,
+      "learning_rate": 4.4554850647634154e-05,
+      "loss": 1.3247,
+      "step": 20609
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8654121160507202,
+      "learning_rate": 4.4554586307163634e-05,
+      "loss": 1.429,
+      "step": 20610
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.8136017322540283,
+      "learning_rate": 4.45543219666931e-05,
+      "loss": 1.2301,
+      "step": 20611
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4297847747802734,
+      "learning_rate": 4.455405762622258e-05,
+      "loss": 1.5097,
+      "step": 20612
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.5345640182495117,
+      "learning_rate": 4.455379328575205e-05,
+      "loss": 1.545,
+      "step": 20613
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9703487157821655,
+      "learning_rate": 4.455352894528153e-05,
+      "loss": 1.3018,
+      "step": 20614
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0641400814056396,
+      "learning_rate": 4.4553264604811e-05,
+      "loss": 1.2815,
+      "step": 20615
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8287808895111084,
+      "learning_rate": 4.4553000264340475e-05,
+      "loss": 1.3293,
+      "step": 20616
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8772882223129272,
+      "learning_rate": 4.455273592386995e-05,
+      "loss": 1.3097,
+      "step": 20617
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.143531560897827,
+      "learning_rate": 4.4552471583399416e-05,
+      "loss": 1.4301,
+      "step": 20618
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.80054771900177,
+      "learning_rate": 4.4552207242928896e-05,
+      "loss": 1.1875,
+      "step": 20619
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.558145761489868,
+      "learning_rate": 4.455194290245837e-05,
+      "loss": 1.2,
+      "step": 20620
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7289187908172607,
+      "learning_rate": 4.455167856198784e-05,
+      "loss": 1.0314,
+      "step": 20621
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.618072509765625,
+      "learning_rate": 4.455141422151732e-05,
+      "loss": 1.2296,
+      "step": 20622
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.817734718322754,
+      "learning_rate": 4.455114988104679e-05,
+      "loss": 1.335,
+      "step": 20623
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0690395832061768,
+      "learning_rate": 4.4550885540576264e-05,
+      "loss": 1.3438,
+      "step": 20624
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8154698610305786,
+      "learning_rate": 4.4550621200105744e-05,
+      "loss": 1.2707,
+      "step": 20625
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.897351861000061,
+      "learning_rate": 4.455035685963521e-05,
+      "loss": 1.2187,
+      "step": 20626
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.5267490148544312,
+      "learning_rate": 4.4550092519164685e-05,
+      "loss": 1.5653,
+      "step": 20627
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.351309299468994,
+      "learning_rate": 4.454982817869416e-05,
+      "loss": 1.3567,
+      "step": 20628
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.123757839202881,
+      "learning_rate": 4.454956383822363e-05,
+      "loss": 1.2649,
+      "step": 20629
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7433013916015625,
+      "learning_rate": 4.454929949775311e-05,
+      "loss": 1.1918,
+      "step": 20630
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9650994539260864,
+      "learning_rate": 4.454903515728258e-05,
+      "loss": 1.4041,
+      "step": 20631
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7774620056152344,
+      "learning_rate": 4.454877081681206e-05,
+      "loss": 1.2293,
+      "step": 20632
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1353936195373535,
+      "learning_rate": 4.4548506476341526e-05,
+      "loss": 1.2475,
+      "step": 20633
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0812528133392334,
+      "learning_rate": 4.4548242135871006e-05,
+      "loss": 1.3124,
+      "step": 20634
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.143996000289917,
+      "learning_rate": 4.454797779540048e-05,
+      "loss": 1.335,
+      "step": 20635
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.256079912185669,
+      "learning_rate": 4.4547713454929954e-05,
+      "loss": 1.4628,
+      "step": 20636
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.691925048828125,
+      "learning_rate": 4.454744911445943e-05,
+      "loss": 1.5235,
+      "step": 20637
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.5048539638519287,
+      "learning_rate": 4.45471847739889e-05,
+      "loss": 1.5103,
+      "step": 20638
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0696640014648438,
+      "learning_rate": 4.4546920433518374e-05,
+      "loss": 1.2145,
+      "step": 20639
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9628123044967651,
+      "learning_rate": 4.454665609304785e-05,
+      "loss": 1.1997,
+      "step": 20640
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9451417922973633,
+      "learning_rate": 4.454639175257732e-05,
+      "loss": 1.1526,
+      "step": 20641
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.158843755722046,
+      "learning_rate": 4.4546127412106795e-05,
+      "loss": 1.2044,
+      "step": 20642
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.251147747039795,
+      "learning_rate": 4.454586307163627e-05,
+      "loss": 1.1795,
+      "step": 20643
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0338332653045654,
+      "learning_rate": 4.454559873116574e-05,
+      "loss": 1.2308,
+      "step": 20644
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9108086824417114,
+      "learning_rate": 4.454533439069522e-05,
+      "loss": 1.2013,
+      "step": 20645
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.3985580205917358,
+      "learning_rate": 4.454507005022469e-05,
+      "loss": 1.0914,
+      "step": 20646
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.083941698074341,
+      "learning_rate": 4.454480570975417e-05,
+      "loss": 1.3552,
+      "step": 20647
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.157375335693359,
+      "learning_rate": 4.4544541369283636e-05,
+      "loss": 1.6092,
+      "step": 20648
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0692050457000732,
+      "learning_rate": 4.454427702881312e-05,
+      "loss": 1.3714,
+      "step": 20649
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.475956916809082,
+      "learning_rate": 4.454401268834259e-05,
+      "loss": 1.296,
+      "step": 20650
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.5311986207962036,
+      "learning_rate": 4.4543748347872064e-05,
+      "loss": 1.6984,
+      "step": 20651
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.373898983001709,
+      "learning_rate": 4.454348400740154e-05,
+      "loss": 1.2498,
+      "step": 20652
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.029930353164673,
+      "learning_rate": 4.4543219666931004e-05,
+      "loss": 1.2365,
+      "step": 20653
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.6714634895324707,
+      "learning_rate": 4.4542955326460485e-05,
+      "loss": 1.1155,
+      "step": 20654
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6865919828414917,
+      "learning_rate": 4.454269098598996e-05,
+      "loss": 1.2626,
+      "step": 20655
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.832829475402832,
+      "learning_rate": 4.454242664551943e-05,
+      "loss": 1.2482,
+      "step": 20656
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7281519174575806,
+      "learning_rate": 4.4542162305048905e-05,
+      "loss": 1.1735,
+      "step": 20657
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1641976833343506,
+      "learning_rate": 4.454189796457838e-05,
+      "loss": 1.2668,
+      "step": 20658
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7906113862991333,
+      "learning_rate": 4.454163362410785e-05,
+      "loss": 1.1434,
+      "step": 20659
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.679622769355774,
+      "learning_rate": 4.454136928363733e-05,
+      "loss": 1.3208,
+      "step": 20660
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.2186968326568604,
+      "learning_rate": 4.45411049431668e-05,
+      "loss": 1.3618,
+      "step": 20661
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.3806755542755127,
+      "learning_rate": 4.454084060269628e-05,
+      "loss": 1.2738,
+      "step": 20662
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.525911808013916,
+      "learning_rate": 4.454057626222575e-05,
+      "loss": 1.6123,
+      "step": 20663
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.804591178894043,
+      "learning_rate": 4.454031192175522e-05,
+      "loss": 1.3295,
+      "step": 20664
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7533737421035767,
+      "learning_rate": 4.45400475812847e-05,
+      "loss": 1.2529,
+      "step": 20665
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.7108564376831055,
+      "learning_rate": 4.453978324081417e-05,
+      "loss": 1.3546,
+      "step": 20666
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9798356294631958,
+      "learning_rate": 4.453951890034365e-05,
+      "loss": 1.3426,
+      "step": 20667
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.75150728225708,
+      "learning_rate": 4.4539254559873115e-05,
+      "loss": 1.2165,
+      "step": 20668
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.5671730041503906,
+      "learning_rate": 4.4538990219402595e-05,
+      "loss": 1.3189,
+      "step": 20669
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9712822437286377,
+      "learning_rate": 4.453872587893207e-05,
+      "loss": 1.2089,
+      "step": 20670
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.648996114730835,
+      "learning_rate": 4.453846153846154e-05,
+      "loss": 1.4434,
+      "step": 20671
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7618224620819092,
+      "learning_rate": 4.4538197197991016e-05,
+      "loss": 1.2277,
+      "step": 20672
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.101454496383667,
+      "learning_rate": 4.453793285752049e-05,
+      "loss": 1.264,
+      "step": 20673
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0123257637023926,
+      "learning_rate": 4.453766851704996e-05,
+      "loss": 1.3869,
+      "step": 20674
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.319652795791626,
+      "learning_rate": 4.4537404176579436e-05,
+      "loss": 1.2662,
+      "step": 20675
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4768314361572266,
+      "learning_rate": 4.453713983610891e-05,
+      "loss": 1.4428,
+      "step": 20676
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6988775730133057,
+      "learning_rate": 4.4536875495638383e-05,
+      "loss": 1.1985,
+      "step": 20677
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0587196350097656,
+      "learning_rate": 4.453661115516786e-05,
+      "loss": 1.502,
+      "step": 20678
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.985281229019165,
+      "learning_rate": 4.453634681469733e-05,
+      "loss": 1.4174,
+      "step": 20679
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9861087799072266,
+      "learning_rate": 4.453608247422681e-05,
+      "loss": 1.4502,
+      "step": 20680
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8741333484649658,
+      "learning_rate": 4.453581813375628e-05,
+      "loss": 1.1693,
+      "step": 20681
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4255874156951904,
+      "learning_rate": 4.453555379328576e-05,
+      "loss": 1.4205,
+      "step": 20682
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.981734275817871,
+      "learning_rate": 4.4535289452815225e-05,
+      "loss": 1.6634,
+      "step": 20683
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.8906809091567993,
+      "learning_rate": 4.4535025112344705e-05,
+      "loss": 1.3506,
+      "step": 20684
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.517733573913574,
+      "learning_rate": 4.453476077187418e-05,
+      "loss": 1.4966,
+      "step": 20685
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.550524950027466,
+      "learning_rate": 4.453449643140365e-05,
+      "loss": 1.2697,
+      "step": 20686
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.292757511138916,
+      "learning_rate": 4.4534232090933126e-05,
+      "loss": 1.425,
+      "step": 20687
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.15840744972229,
+      "learning_rate": 4.45339677504626e-05,
+      "loss": 1.2566,
+      "step": 20688
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0063204765319824,
+      "learning_rate": 4.453370340999207e-05,
+      "loss": 1.2343,
+      "step": 20689
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.973865270614624,
+      "learning_rate": 4.453343906952155e-05,
+      "loss": 1.2487,
+      "step": 20690
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9499001502990723,
+      "learning_rate": 4.453317472905102e-05,
+      "loss": 1.2624,
+      "step": 20691
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9493117332458496,
+      "learning_rate": 4.4532910388580494e-05,
+      "loss": 1.38,
+      "step": 20692
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9548043012619019,
+      "learning_rate": 4.453264604810997e-05,
+      "loss": 1.3583,
+      "step": 20693
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0315632820129395,
+      "learning_rate": 4.453238170763944e-05,
+      "loss": 1.1825,
+      "step": 20694
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.6661484241485596,
+      "learning_rate": 4.453211736716892e-05,
+      "loss": 1.4737,
+      "step": 20695
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.3647191524505615,
+      "learning_rate": 4.453185302669839e-05,
+      "loss": 1.3297,
+      "step": 20696
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.3783950805664062,
+      "learning_rate": 4.453158868622787e-05,
+      "loss": 1.3458,
+      "step": 20697
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.004403829574585,
+      "learning_rate": 4.4531324345757335e-05,
+      "loss": 2.2338,
+      "step": 20698
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7472119331359863,
+      "learning_rate": 4.453106000528681e-05,
+      "loss": 1.1672,
+      "step": 20699
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.045119285583496,
+      "learning_rate": 4.453079566481628e-05,
+      "loss": 1.3659,
+      "step": 20700
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6352845430374146,
+      "learning_rate": 4.4530531324345756e-05,
+      "loss": 1.1535,
+      "step": 20701
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.823254942893982,
+      "learning_rate": 4.4530266983875236e-05,
+      "loss": 1.0939,
+      "step": 20702
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.317453384399414,
+      "learning_rate": 4.45300026434047e-05,
+      "loss": 1.3197,
+      "step": 20703
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.937043309211731,
+      "learning_rate": 4.4529738302934183e-05,
+      "loss": 1.3346,
+      "step": 20704
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.869731068611145,
+      "learning_rate": 4.452947396246365e-05,
+      "loss": 1.1936,
+      "step": 20705
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.4943747520446777,
+      "learning_rate": 4.452920962199313e-05,
+      "loss": 1.376,
+      "step": 20706
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.4662222862243652,
+      "learning_rate": 4.4528945281522604e-05,
+      "loss": 1.2088,
+      "step": 20707
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.827458143234253,
+      "learning_rate": 4.452868094105208e-05,
+      "loss": 1.2782,
+      "step": 20708
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6322383880615234,
+      "learning_rate": 4.452841660058155e-05,
+      "loss": 1.9506,
+      "step": 20709
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.461724877357483,
+      "learning_rate": 4.4528152260111025e-05,
+      "loss": 1.3668,
+      "step": 20710
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.8906474113464355,
+      "learning_rate": 4.45278879196405e-05,
+      "loss": 1.2728,
+      "step": 20711
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.326464891433716,
+      "learning_rate": 4.452762357916997e-05,
+      "loss": 1.3925,
+      "step": 20712
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0852179527282715,
+      "learning_rate": 4.4527359238699446e-05,
+      "loss": 1.1504,
+      "step": 20713
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8099278211593628,
+      "learning_rate": 4.452709489822892e-05,
+      "loss": 1.3023,
+      "step": 20714
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.9446122646331787,
+      "learning_rate": 4.452683055775839e-05,
+      "loss": 1.3863,
+      "step": 20715
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.763100266456604,
+      "learning_rate": 4.4526566217287866e-05,
+      "loss": 1.4971,
+      "step": 20716
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1195662021636963,
+      "learning_rate": 4.452630187681735e-05,
+      "loss": 1.2547,
+      "step": 20717
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8952544927597046,
+      "learning_rate": 4.4526037536346813e-05,
+      "loss": 1.2378,
+      "step": 20718
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8344776630401611,
+      "learning_rate": 4.4525773195876294e-05,
+      "loss": 1.3171,
+      "step": 20719
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9954078197479248,
+      "learning_rate": 4.452550885540576e-05,
+      "loss": 1.4458,
+      "step": 20720
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1185038089752197,
+      "learning_rate": 4.452524451493524e-05,
+      "loss": 1.3434,
+      "step": 20721
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.897018313407898,
+      "learning_rate": 4.4524980174464715e-05,
+      "loss": 1.22,
+      "step": 20722
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1198365688323975,
+      "learning_rate": 4.452471583399419e-05,
+      "loss": 1.3704,
+      "step": 20723
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.849697232246399,
+      "learning_rate": 4.452445149352366e-05,
+      "loss": 1.1907,
+      "step": 20724
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.185957908630371,
+      "learning_rate": 4.452418715305313e-05,
+      "loss": 1.2969,
+      "step": 20725
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.7038421630859375,
+      "learning_rate": 4.452392281258261e-05,
+      "loss": 1.4487,
+      "step": 20726
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9643807411193848,
+      "learning_rate": 4.452365847211208e-05,
+      "loss": 1.2117,
+      "step": 20727
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9527060985565186,
+      "learning_rate": 4.4523394131641556e-05,
+      "loss": 1.2676,
+      "step": 20728
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7906430959701538,
+      "learning_rate": 4.452312979117103e-05,
+      "loss": 1.2427,
+      "step": 20729
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.914886236190796,
+      "learning_rate": 4.45228654507005e-05,
+      "loss": 1.2884,
+      "step": 20730
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.17561936378479,
+      "learning_rate": 4.452260111022998e-05,
+      "loss": 1.2905,
+      "step": 20731
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.5695242881774902,
+      "learning_rate": 4.452233676975946e-05,
+      "loss": 1.0804,
+      "step": 20732
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.210477113723755,
+      "learning_rate": 4.4522072429288924e-05,
+      "loss": 1.2805,
+      "step": 20733
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1840083599090576,
+      "learning_rate": 4.4521808088818404e-05,
+      "loss": 1.3149,
+      "step": 20734
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.4515488147735596,
+      "learning_rate": 4.452154374834787e-05,
+      "loss": 1.3801,
+      "step": 20735
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8621416091918945,
+      "learning_rate": 4.4521279407877345e-05,
+      "loss": 1.3279,
+      "step": 20736
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0902576446533203,
+      "learning_rate": 4.4521015067406825e-05,
+      "loss": 1.2207,
+      "step": 20737
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.8507936000823975,
+      "learning_rate": 4.452075072693629e-05,
+      "loss": 1.3163,
+      "step": 20738
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0720906257629395,
+      "learning_rate": 4.452048638646577e-05,
+      "loss": 1.3802,
+      "step": 20739
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7391667366027832,
+      "learning_rate": 4.452022204599524e-05,
+      "loss": 1.1522,
+      "step": 20740
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.982249140739441,
+      "learning_rate": 4.451995770552472e-05,
+      "loss": 1.2753,
+      "step": 20741
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9677447080612183,
+      "learning_rate": 4.451969336505419e-05,
+      "loss": 1.2647,
+      "step": 20742
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.003340482711792,
+      "learning_rate": 4.4519429024583666e-05,
+      "loss": 1.193,
+      "step": 20743
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1204535961151123,
+      "learning_rate": 4.451916468411314e-05,
+      "loss": 1.4058,
+      "step": 20744
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.077739715576172,
+      "learning_rate": 4.4518900343642613e-05,
+      "loss": 1.3216,
+      "step": 20745
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2483277320861816,
+      "learning_rate": 4.451863600317209e-05,
+      "loss": 1.3604,
+      "step": 20746
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.841535210609436,
+      "learning_rate": 4.451837166270156e-05,
+      "loss": 1.2664,
+      "step": 20747
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5100085735321045,
+      "learning_rate": 4.4518107322231034e-05,
+      "loss": 1.4278,
+      "step": 20748
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.9841127395629883,
+      "learning_rate": 4.451784298176051e-05,
+      "loss": 1.449,
+      "step": 20749
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8588550090789795,
+      "learning_rate": 4.451757864128998e-05,
+      "loss": 1.311,
+      "step": 20750
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.158994674682617,
+      "learning_rate": 4.4517314300819455e-05,
+      "loss": 1.4255,
+      "step": 20751
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.497424364089966,
+      "learning_rate": 4.4517049960348935e-05,
+      "loss": 1.463,
+      "step": 20752
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.46592378616333,
+      "learning_rate": 4.45167856198784e-05,
+      "loss": 1.3495,
+      "step": 20753
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5225772857666016,
+      "learning_rate": 4.451652127940788e-05,
+      "loss": 1.2395,
+      "step": 20754
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8514572381973267,
+      "learning_rate": 4.451625693893735e-05,
+      "loss": 1.2694,
+      "step": 20755
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9674780368804932,
+      "learning_rate": 4.451599259846683e-05,
+      "loss": 1.3153,
+      "step": 20756
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1323964595794678,
+      "learning_rate": 4.45157282579963e-05,
+      "loss": 1.2897,
+      "step": 20757
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.9479663372039795,
+      "learning_rate": 4.451546391752578e-05,
+      "loss": 1.4051,
+      "step": 20758
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1706793308258057,
+      "learning_rate": 4.451519957705525e-05,
+      "loss": 1.5951,
+      "step": 20759
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7783420085906982,
+      "learning_rate": 4.4514935236584724e-05,
+      "loss": 1.1808,
+      "step": 20760
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.851465106010437,
+      "learning_rate": 4.45146708961142e-05,
+      "loss": 1.2437,
+      "step": 20761
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1757771968841553,
+      "learning_rate": 4.451440655564367e-05,
+      "loss": 1.3524,
+      "step": 20762
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1881303787231445,
+      "learning_rate": 4.4514142215173144e-05,
+      "loss": 1.3074,
+      "step": 20763
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6845005750656128,
+      "learning_rate": 4.451387787470262e-05,
+      "loss": 1.2885,
+      "step": 20764
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.3588266372680664,
+      "learning_rate": 4.451361353423209e-05,
+      "loss": 1.5466,
+      "step": 20765
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.0492663383483887,
+      "learning_rate": 4.4513349193761565e-05,
+      "loss": 1.3829,
+      "step": 20766
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0357508659362793,
+      "learning_rate": 4.4513084853291046e-05,
+      "loss": 1.1829,
+      "step": 20767
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.4635956287384033,
+      "learning_rate": 4.451282051282051e-05,
+      "loss": 1.4997,
+      "step": 20768
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.732236385345459,
+      "learning_rate": 4.451255617234999e-05,
+      "loss": 1.2435,
+      "step": 20769
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.241471290588379,
+      "learning_rate": 4.451229183187946e-05,
+      "loss": 1.2385,
+      "step": 20770
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0947024822235107,
+      "learning_rate": 4.451202749140893e-05,
+      "loss": 1.3811,
+      "step": 20771
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9809085130691528,
+      "learning_rate": 4.451176315093841e-05,
+      "loss": 1.3661,
+      "step": 20772
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.5428799390792847,
+      "learning_rate": 4.451149881046788e-05,
+      "loss": 1.2575,
+      "step": 20773
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.4609360694885254,
+      "learning_rate": 4.451123446999736e-05,
+      "loss": 1.2443,
+      "step": 20774
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5565946102142334,
+      "learning_rate": 4.451097012952683e-05,
+      "loss": 1.3302,
+      "step": 20775
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0458552837371826,
+      "learning_rate": 4.451070578905631e-05,
+      "loss": 1.3818,
+      "step": 20776
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0985829830169678,
+      "learning_rate": 4.451044144858578e-05,
+      "loss": 1.299,
+      "step": 20777
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0808873176574707,
+      "learning_rate": 4.4510177108115255e-05,
+      "loss": 1.3797,
+      "step": 20778
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.737947702407837,
+      "learning_rate": 4.450991276764473e-05,
+      "loss": 1.1826,
+      "step": 20779
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.579899549484253,
+      "learning_rate": 4.45096484271742e-05,
+      "loss": 1.3647,
+      "step": 20780
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.217618942260742,
+      "learning_rate": 4.4509384086703676e-05,
+      "loss": 1.2441,
+      "step": 20781
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9521507024765015,
+      "learning_rate": 4.450911974623315e-05,
+      "loss": 1.2607,
+      "step": 20782
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8625236749649048,
+      "learning_rate": 4.450885540576262e-05,
+      "loss": 1.2731,
+      "step": 20783
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8105926513671875,
+      "learning_rate": 4.4508591065292096e-05,
+      "loss": 1.2764,
+      "step": 20784
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9797009229660034,
+      "learning_rate": 4.450832672482157e-05,
+      "loss": 1.336,
+      "step": 20785
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.502032995223999,
+      "learning_rate": 4.450806238435104e-05,
+      "loss": 1.4937,
+      "step": 20786
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9007726907730103,
+      "learning_rate": 4.4507798043880524e-05,
+      "loss": 1.3875,
+      "step": 20787
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.033043146133423,
+      "learning_rate": 4.450753370340999e-05,
+      "loss": 1.4029,
+      "step": 20788
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8853776454925537,
+      "learning_rate": 4.450726936293947e-05,
+      "loss": 1.9192,
+      "step": 20789
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8311538696289062,
+      "learning_rate": 4.450700502246894e-05,
+      "loss": 1.5637,
+      "step": 20790
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2714061737060547,
+      "learning_rate": 4.450674068199842e-05,
+      "loss": 1.3548,
+      "step": 20791
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.885399103164673,
+      "learning_rate": 4.450647634152789e-05,
+      "loss": 1.4971,
+      "step": 20792
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1597328186035156,
+      "learning_rate": 4.4506212001057365e-05,
+      "loss": 1.4962,
+      "step": 20793
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2966644763946533,
+      "learning_rate": 4.450594766058684e-05,
+      "loss": 1.4226,
+      "step": 20794
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0810580253601074,
+      "learning_rate": 4.450568332011631e-05,
+      "loss": 1.5545,
+      "step": 20795
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.478444814682007,
+      "learning_rate": 4.4505418979645786e-05,
+      "loss": 1.2948,
+      "step": 20796
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1852595806121826,
+      "learning_rate": 4.450515463917526e-05,
+      "loss": 1.2996,
+      "step": 20797
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.021165370941162,
+      "learning_rate": 4.450489029870473e-05,
+      "loss": 1.2132,
+      "step": 20798
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.059046745300293,
+      "learning_rate": 4.4504625958234207e-05,
+      "loss": 1.3387,
+      "step": 20799
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.212585926055908,
+      "learning_rate": 4.450436161776368e-05,
+      "loss": 1.3715,
+      "step": 20800
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.6498701572418213,
+      "learning_rate": 4.4504097277293154e-05,
+      "loss": 1.3738,
+      "step": 20801
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7870821952819824,
+      "learning_rate": 4.4503832936822634e-05,
+      "loss": 1.7611,
+      "step": 20802
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1058976650238037,
+      "learning_rate": 4.45035685963521e-05,
+      "loss": 2.2654,
+      "step": 20803
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.029554605484009,
+      "learning_rate": 4.450330425588158e-05,
+      "loss": 1.2694,
+      "step": 20804
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.108694553375244,
+      "learning_rate": 4.450303991541105e-05,
+      "loss": 1.3498,
+      "step": 20805
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9265674352645874,
+      "learning_rate": 4.450277557494053e-05,
+      "loss": 1.0677,
+      "step": 20806
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.141691207885742,
+      "learning_rate": 4.450251123447e-05,
+      "loss": 1.2929,
+      "step": 20807
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.009371042251587,
+      "learning_rate": 4.450224689399947e-05,
+      "loss": 1.2723,
+      "step": 20808
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0495150089263916,
+      "learning_rate": 4.450198255352895e-05,
+      "loss": 1.2372,
+      "step": 20809
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7837955951690674,
+      "learning_rate": 4.4501718213058416e-05,
+      "loss": 1.2557,
+      "step": 20810
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8191050291061401,
+      "learning_rate": 4.4501453872587896e-05,
+      "loss": 1.3304,
+      "step": 20811
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8428996801376343,
+      "learning_rate": 4.450118953211737e-05,
+      "loss": 1.2968,
+      "step": 20812
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8698573112487793,
+      "learning_rate": 4.450092519164684e-05,
+      "loss": 1.343,
+      "step": 20813
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.023461103439331,
+      "learning_rate": 4.450066085117632e-05,
+      "loss": 1.1848,
+      "step": 20814
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.1465704441070557,
+      "learning_rate": 4.450039651070579e-05,
+      "loss": 1.3128,
+      "step": 20815
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7869302034378052,
+      "learning_rate": 4.4500132170235264e-05,
+      "loss": 1.2256,
+      "step": 20816
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.3197031021118164,
+      "learning_rate": 4.449986782976474e-05,
+      "loss": 1.3601,
+      "step": 20817
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.154280424118042,
+      "learning_rate": 4.449960348929421e-05,
+      "loss": 1.4216,
+      "step": 20818
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.657233715057373,
+      "learning_rate": 4.4499339148823685e-05,
+      "loss": 1.276,
+      "step": 20819
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.87929105758667,
+      "learning_rate": 4.449907480835316e-05,
+      "loss": 1.2702,
+      "step": 20820
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9196548461914062,
+      "learning_rate": 4.449881046788263e-05,
+      "loss": 1.2547,
+      "step": 20821
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1443629264831543,
+      "learning_rate": 4.449854612741211e-05,
+      "loss": 1.2319,
+      "step": 20822
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6434632539749146,
+      "learning_rate": 4.449828178694158e-05,
+      "loss": 1.3066,
+      "step": 20823
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5478174686431885,
+      "learning_rate": 4.449801744647106e-05,
+      "loss": 1.5252,
+      "step": 20824
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8677475452423096,
+      "learning_rate": 4.4497753106000526e-05,
+      "loss": 1.2398,
+      "step": 20825
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9742158651351929,
+      "learning_rate": 4.4497488765530007e-05,
+      "loss": 1.336,
+      "step": 20826
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7020621299743652,
+      "learning_rate": 4.449722442505948e-05,
+      "loss": 1.3721,
+      "step": 20827
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6912816762924194,
+      "learning_rate": 4.4496960084588954e-05,
+      "loss": 1.1297,
+      "step": 20828
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.661064624786377,
+      "learning_rate": 4.449669574411843e-05,
+      "loss": 1.5262,
+      "step": 20829
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6338237524032593,
+      "learning_rate": 4.44964314036479e-05,
+      "loss": 1.0695,
+      "step": 20830
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.134364128112793,
+      "learning_rate": 4.4496167063177374e-05,
+      "loss": 1.279,
+      "step": 20831
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8938567638397217,
+      "learning_rate": 4.449590272270685e-05,
+      "loss": 1.213,
+      "step": 20832
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9530328512191772,
+      "learning_rate": 4.449563838223632e-05,
+      "loss": 1.221,
+      "step": 20833
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9156755208969116,
+      "learning_rate": 4.4495374041765795e-05,
+      "loss": 1.3345,
+      "step": 20834
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.5675755739212036,
+      "learning_rate": 4.449510970129527e-05,
+      "loss": 1.2076,
+      "step": 20835
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.689976215362549,
+      "learning_rate": 4.449484536082474e-05,
+      "loss": 1.2984,
+      "step": 20836
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.017592668533325,
+      "learning_rate": 4.449458102035422e-05,
+      "loss": 1.1543,
+      "step": 20837
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.4873008728027344,
+      "learning_rate": 4.449431667988369e-05,
+      "loss": 1.3472,
+      "step": 20838
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.819273829460144,
+      "learning_rate": 4.449405233941317e-05,
+      "loss": 1.1024,
+      "step": 20839
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.902098536491394,
+      "learning_rate": 4.4493787998942637e-05,
+      "loss": 1.3159,
+      "step": 20840
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.016017436981201,
+      "learning_rate": 4.449352365847212e-05,
+      "loss": 1.2884,
+      "step": 20841
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.070535182952881,
+      "learning_rate": 4.449325931800159e-05,
+      "loss": 1.2882,
+      "step": 20842
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5255136489868164,
+      "learning_rate": 4.449299497753106e-05,
+      "loss": 1.1857,
+      "step": 20843
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.758023977279663,
+      "learning_rate": 4.449273063706054e-05,
+      "loss": 1.3682,
+      "step": 20844
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.133223056793213,
+      "learning_rate": 4.4492466296590004e-05,
+      "loss": 1.4021,
+      "step": 20845
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7449275255203247,
+      "learning_rate": 4.4492201956119485e-05,
+      "loss": 1.3165,
+      "step": 20846
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.937483787536621,
+      "learning_rate": 4.449193761564896e-05,
+      "loss": 1.2464,
+      "step": 20847
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8619619607925415,
+      "learning_rate": 4.449167327517843e-05,
+      "loss": 1.1225,
+      "step": 20848
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.0027012825012207,
+      "learning_rate": 4.4491408934707905e-05,
+      "loss": 1.4108,
+      "step": 20849
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.286404848098755,
+      "learning_rate": 4.449114459423738e-05,
+      "loss": 1.364,
+      "step": 20850
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6974284648895264,
+      "learning_rate": 4.449088025376685e-05,
+      "loss": 1.1816,
+      "step": 20851
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9279588460922241,
+      "learning_rate": 4.449061591329633e-05,
+      "loss": 1.2892,
+      "step": 20852
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2110066413879395,
+      "learning_rate": 4.44903515728258e-05,
+      "loss": 1.302,
+      "step": 20853
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8765867948532104,
+      "learning_rate": 4.449008723235527e-05,
+      "loss": 1.1725,
+      "step": 20854
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1419694423675537,
+      "learning_rate": 4.448982289188475e-05,
+      "loss": 1.293,
+      "step": 20855
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1928365230560303,
+      "learning_rate": 4.448955855141422e-05,
+      "loss": 1.1831,
+      "step": 20856
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.101339817047119,
+      "learning_rate": 4.44892942109437e-05,
+      "loss": 1.3888,
+      "step": 20857
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2189297676086426,
+      "learning_rate": 4.448902987047317e-05,
+      "loss": 1.2603,
+      "step": 20858
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0933940410614014,
+      "learning_rate": 4.448876553000265e-05,
+      "loss": 1.4811,
+      "step": 20859
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8077832460403442,
+      "learning_rate": 4.4488501189532115e-05,
+      "loss": 1.3565,
+      "step": 20860
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.191023826599121,
+      "learning_rate": 4.4488236849061595e-05,
+      "loss": 1.195,
+      "step": 20861
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9558721780776978,
+      "learning_rate": 4.448797250859107e-05,
+      "loss": 1.2435,
+      "step": 20862
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5437517166137695,
+      "learning_rate": 4.448770816812054e-05,
+      "loss": 1.508,
+      "step": 20863
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7924377918243408,
+      "learning_rate": 4.4487443827650016e-05,
+      "loss": 1.3246,
+      "step": 20864
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.761043906211853,
+      "learning_rate": 4.448717948717949e-05,
+      "loss": 1.1841,
+      "step": 20865
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8101478815078735,
+      "learning_rate": 4.448691514670896e-05,
+      "loss": 1.3089,
+      "step": 20866
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.756727933883667,
+      "learning_rate": 4.4486650806238437e-05,
+      "loss": 1.1167,
+      "step": 20867
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.910296082496643,
+      "learning_rate": 4.448638646576791e-05,
+      "loss": 1.2731,
+      "step": 20868
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.84812331199646,
+      "learning_rate": 4.4486122125297384e-05,
+      "loss": 1.3984,
+      "step": 20869
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1142685413360596,
+      "learning_rate": 4.448585778482686e-05,
+      "loss": 1.366,
+      "step": 20870
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8134527206420898,
+      "learning_rate": 4.448559344435633e-05,
+      "loss": 1.2641,
+      "step": 20871
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.773422122001648,
+      "learning_rate": 4.448532910388581e-05,
+      "loss": 1.2615,
+      "step": 20872
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.048231840133667,
+      "learning_rate": 4.448506476341528e-05,
+      "loss": 1.4685,
+      "step": 20873
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8548297882080078,
+      "learning_rate": 4.448480042294476e-05,
+      "loss": 1.3464,
+      "step": 20874
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.381324291229248,
+      "learning_rate": 4.4484536082474225e-05,
+      "loss": 1.235,
+      "step": 20875
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.227440118789673,
+      "learning_rate": 4.4484271742003705e-05,
+      "loss": 1.3153,
+      "step": 20876
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.302457571029663,
+      "learning_rate": 4.448400740153318e-05,
+      "loss": 1.4021,
+      "step": 20877
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8805184364318848,
+      "learning_rate": 4.448374306106265e-05,
+      "loss": 1.2696,
+      "step": 20878
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.7621915340423584,
+      "learning_rate": 4.4483478720592126e-05,
+      "loss": 1.3848,
+      "step": 20879
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.526620864868164,
+      "learning_rate": 4.448321438012159e-05,
+      "loss": 1.3826,
+      "step": 20880
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1157798767089844,
+      "learning_rate": 4.448295003965107e-05,
+      "loss": 1.2475,
+      "step": 20881
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.2609360218048096,
+      "learning_rate": 4.448268569918055e-05,
+      "loss": 1.2753,
+      "step": 20882
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0945167541503906,
+      "learning_rate": 4.448242135871002e-05,
+      "loss": 1.3432,
+      "step": 20883
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1536238193511963,
+      "learning_rate": 4.4482157018239494e-05,
+      "loss": 1.1943,
+      "step": 20884
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5088753700256348,
+      "learning_rate": 4.448189267776897e-05,
+      "loss": 1.3366,
+      "step": 20885
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9017831087112427,
+      "learning_rate": 4.448162833729844e-05,
+      "loss": 1.3082,
+      "step": 20886
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0528295040130615,
+      "learning_rate": 4.448136399682792e-05,
+      "loss": 1.3935,
+      "step": 20887
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8515232801437378,
+      "learning_rate": 4.448109965635739e-05,
+      "loss": 1.2469,
+      "step": 20888
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.071439504623413,
+      "learning_rate": 4.448083531588686e-05,
+      "loss": 1.3074,
+      "step": 20889
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2826743125915527,
+      "learning_rate": 4.4480570975416335e-05,
+      "loss": 1.2706,
+      "step": 20890
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.123302459716797,
+      "learning_rate": 4.448030663494581e-05,
+      "loss": 1.2987,
+      "step": 20891
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.087448835372925,
+      "learning_rate": 4.448004229447529e-05,
+      "loss": 1.2867,
+      "step": 20892
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.603094220161438,
+      "learning_rate": 4.4479777954004756e-05,
+      "loss": 1.2067,
+      "step": 20893
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.992651343345642,
+      "learning_rate": 4.4479513613534236e-05,
+      "loss": 1.3124,
+      "step": 20894
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7498620748519897,
+      "learning_rate": 4.44792492730637e-05,
+      "loss": 1.3548,
+      "step": 20895
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1314618587493896,
+      "learning_rate": 4.4478984932593184e-05,
+      "loss": 1.1998,
+      "step": 20896
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.807607650756836,
+      "learning_rate": 4.447872059212266e-05,
+      "loss": 1.2799,
+      "step": 20897
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9769091606140137,
+      "learning_rate": 4.447845625165213e-05,
+      "loss": 1.4302,
+      "step": 20898
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0092153549194336,
+      "learning_rate": 4.4478191911181604e-05,
+      "loss": 1.2833,
+      "step": 20899
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7762324810028076,
+      "learning_rate": 4.447792757071108e-05,
+      "loss": 1.1716,
+      "step": 20900
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2725026607513428,
+      "learning_rate": 4.447766323024055e-05,
+      "loss": 1.4153,
+      "step": 20901
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2727608680725098,
+      "learning_rate": 4.4477398889770025e-05,
+      "loss": 1.4276,
+      "step": 20902
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2273545265197754,
+      "learning_rate": 4.44771345492995e-05,
+      "loss": 1.4305,
+      "step": 20903
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1914894580841064,
+      "learning_rate": 4.447687020882897e-05,
+      "loss": 1.3449,
+      "step": 20904
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.978711485862732,
+      "learning_rate": 4.4476605868358446e-05,
+      "loss": 1.2463,
+      "step": 20905
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.137964963912964,
+      "learning_rate": 4.447634152788792e-05,
+      "loss": 1.383,
+      "step": 20906
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.4821553230285645,
+      "learning_rate": 4.44760771874174e-05,
+      "loss": 1.4084,
+      "step": 20907
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.240220069885254,
+      "learning_rate": 4.4475812846946866e-05,
+      "loss": 1.2222,
+      "step": 20908
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.007640838623047,
+      "learning_rate": 4.447554850647635e-05,
+      "loss": 1.2279,
+      "step": 20909
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.040541172027588,
+      "learning_rate": 4.4475284166005814e-05,
+      "loss": 1.2899,
+      "step": 20910
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.5667951107025146,
+      "learning_rate": 4.4475019825535294e-05,
+      "loss": 1.4861,
+      "step": 20911
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.6096203327178955,
+      "learning_rate": 4.447475548506477e-05,
+      "loss": 1.3539,
+      "step": 20912
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.105787515640259,
+      "learning_rate": 4.447449114459424e-05,
+      "loss": 1.2633,
+      "step": 20913
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8770275115966797,
+      "learning_rate": 4.4474226804123715e-05,
+      "loss": 1.1752,
+      "step": 20914
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1454780101776123,
+      "learning_rate": 4.447396246365319e-05,
+      "loss": 1.2814,
+      "step": 20915
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0073070526123047,
+      "learning_rate": 4.447369812318266e-05,
+      "loss": 1.189,
+      "step": 20916
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1565065383911133,
+      "learning_rate": 4.4473433782712135e-05,
+      "loss": 1.3621,
+      "step": 20917
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0529236793518066,
+      "learning_rate": 4.447316944224161e-05,
+      "loss": 1.3657,
+      "step": 20918
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9142403602600098,
+      "learning_rate": 4.447290510177108e-05,
+      "loss": 1.258,
+      "step": 20919
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0497870445251465,
+      "learning_rate": 4.4472640761300556e-05,
+      "loss": 1.4136,
+      "step": 20920
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.6243312358856201,
+      "learning_rate": 4.447237642083003e-05,
+      "loss": 1.1308,
+      "step": 20921
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8090555667877197,
+      "learning_rate": 4.447211208035951e-05,
+      "loss": 1.1996,
+      "step": 20922
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.031601667404175,
+      "learning_rate": 4.447184773988898e-05,
+      "loss": 1.235,
+      "step": 20923
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.151675224304199,
+      "learning_rate": 4.447158339941846e-05,
+      "loss": 1.4727,
+      "step": 20924
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.946933627128601,
+      "learning_rate": 4.4471319058947924e-05,
+      "loss": 1.3137,
+      "step": 20925
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1830856800079346,
+      "learning_rate": 4.44710547184774e-05,
+      "loss": 1.2824,
+      "step": 20926
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.548513412475586,
+      "learning_rate": 4.447079037800688e-05,
+      "loss": 1.3683,
+      "step": 20927
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9750256538391113,
+      "learning_rate": 4.4470526037536345e-05,
+      "loss": 1.3578,
+      "step": 20928
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.4025018215179443,
+      "learning_rate": 4.4470261697065825e-05,
+      "loss": 1.4366,
+      "step": 20929
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1733853816986084,
+      "learning_rate": 4.446999735659529e-05,
+      "loss": 1.3964,
+      "step": 20930
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8895783424377441,
+      "learning_rate": 4.446973301612477e-05,
+      "loss": 1.2667,
+      "step": 20931
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.680049180984497,
+      "learning_rate": 4.4469468675654246e-05,
+      "loss": 1.2823,
+      "step": 20932
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.738241195678711,
+      "learning_rate": 4.446920433518372e-05,
+      "loss": 1.2551,
+      "step": 20933
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.4623072147369385,
+      "learning_rate": 4.446893999471319e-05,
+      "loss": 1.3962,
+      "step": 20934
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9645271301269531,
+      "learning_rate": 4.4468675654242666e-05,
+      "loss": 1.2592,
+      "step": 20935
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.5148903131484985,
+      "learning_rate": 4.446841131377214e-05,
+      "loss": 1.1208,
+      "step": 20936
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.56192684173584,
+      "learning_rate": 4.4468146973301614e-05,
+      "loss": 1.4162,
+      "step": 20937
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.619919776916504,
+      "learning_rate": 4.446788263283109e-05,
+      "loss": 1.2683,
+      "step": 20938
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2568211555480957,
+      "learning_rate": 4.446761829236056e-05,
+      "loss": 1.4229,
+      "step": 20939
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.106584310531616,
+      "learning_rate": 4.4467353951890034e-05,
+      "loss": 1.3209,
+      "step": 20940
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.348909616470337,
+      "learning_rate": 4.446708961141951e-05,
+      "loss": 1.1616,
+      "step": 20941
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9045192003250122,
+      "learning_rate": 4.446682527094899e-05,
+      "loss": 1.3003,
+      "step": 20942
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.5903135538101196,
+      "learning_rate": 4.4466560930478455e-05,
+      "loss": 1.2118,
+      "step": 20943
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.79108726978302,
+      "learning_rate": 4.4466296590007935e-05,
+      "loss": 1.2484,
+      "step": 20944
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.685206890106201,
+      "learning_rate": 4.44660322495374e-05,
+      "loss": 1.3823,
+      "step": 20945
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.825086236000061,
+      "learning_rate": 4.446576790906688e-05,
+      "loss": 1.2481,
+      "step": 20946
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.379101514816284,
+      "learning_rate": 4.4465503568596356e-05,
+      "loss": 1.265,
+      "step": 20947
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.135617971420288,
+      "learning_rate": 4.446523922812583e-05,
+      "loss": 1.2916,
+      "step": 20948
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8474023342132568,
+      "learning_rate": 4.44649748876553e-05,
+      "loss": 1.1972,
+      "step": 20949
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.545063495635986,
+      "learning_rate": 4.446471054718478e-05,
+      "loss": 1.4548,
+      "step": 20950
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.209686517715454,
+      "learning_rate": 4.446444620671425e-05,
+      "loss": 1.193,
+      "step": 20951
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8667587041854858,
+      "learning_rate": 4.4464181866243724e-05,
+      "loss": 1.326,
+      "step": 20952
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.808411717414856,
+      "learning_rate": 4.44639175257732e-05,
+      "loss": 1.127,
+      "step": 20953
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0305256843566895,
+      "learning_rate": 4.446365318530267e-05,
+      "loss": 1.2183,
+      "step": 20954
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.211486577987671,
+      "learning_rate": 4.4463388844832145e-05,
+      "loss": 1.4112,
+      "step": 20955
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.4636468887329102,
+      "learning_rate": 4.446312450436162e-05,
+      "loss": 1.4676,
+      "step": 20956
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1392061710357666,
+      "learning_rate": 4.44628601638911e-05,
+      "loss": 1.4059,
+      "step": 20957
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0406627655029297,
+      "learning_rate": 4.4462595823420565e-05,
+      "loss": 1.2984,
+      "step": 20958
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7235711812973022,
+      "learning_rate": 4.4462331482950046e-05,
+      "loss": 1.2514,
+      "step": 20959
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9480599164962769,
+      "learning_rate": 4.446206714247951e-05,
+      "loss": 1.3387,
+      "step": 20960
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7474397420883179,
+      "learning_rate": 4.4461802802008986e-05,
+      "loss": 1.2715,
+      "step": 20961
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7494945526123047,
+      "learning_rate": 4.4461538461538466e-05,
+      "loss": 1.1957,
+      "step": 20962
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9520635604858398,
+      "learning_rate": 4.446127412106793e-05,
+      "loss": 1.2345,
+      "step": 20963
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.708965539932251,
+      "learning_rate": 4.4461009780597414e-05,
+      "loss": 1.34,
+      "step": 20964
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9498488903045654,
+      "learning_rate": 4.446074544012688e-05,
+      "loss": 1.2587,
+      "step": 20965
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.111100435256958,
+      "learning_rate": 4.446048109965636e-05,
+      "loss": 1.2438,
+      "step": 20966
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.395297050476074,
+      "learning_rate": 4.4460216759185834e-05,
+      "loss": 1.1294,
+      "step": 20967
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0851686000823975,
+      "learning_rate": 4.445995241871531e-05,
+      "loss": 1.2666,
+      "step": 20968
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7366437911987305,
+      "learning_rate": 4.445968807824478e-05,
+      "loss": 1.2161,
+      "step": 20969
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9305466413497925,
+      "learning_rate": 4.4459423737774255e-05,
+      "loss": 1.2676,
+      "step": 20970
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.013702154159546,
+      "learning_rate": 4.445915939730373e-05,
+      "loss": 1.3967,
+      "step": 20971
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.284188985824585,
+      "learning_rate": 4.44588950568332e-05,
+      "loss": 1.3367,
+      "step": 20972
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9405659437179565,
+      "learning_rate": 4.4458630716362676e-05,
+      "loss": 1.288,
+      "step": 20973
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.8607184886932373,
+      "learning_rate": 4.445836637589215e-05,
+      "loss": 1.624,
+      "step": 20974
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9066567420959473,
+      "learning_rate": 4.445810203542162e-05,
+      "loss": 1.3299,
+      "step": 20975
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.025388240814209,
+      "learning_rate": 4.4457837694951096e-05,
+      "loss": 1.3616,
+      "step": 20976
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9648524522781372,
+      "learning_rate": 4.445757335448058e-05,
+      "loss": 1.1747,
+      "step": 20977
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9461417198181152,
+      "learning_rate": 4.4457309014010044e-05,
+      "loss": 2.2246,
+      "step": 20978
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8565478324890137,
+      "learning_rate": 4.4457044673539524e-05,
+      "loss": 1.1234,
+      "step": 20979
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3400378227233887,
+      "learning_rate": 4.445678033306899e-05,
+      "loss": 1.2995,
+      "step": 20980
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7614343166351318,
+      "learning_rate": 4.445651599259847e-05,
+      "loss": 1.2538,
+      "step": 20981
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.800973892211914,
+      "learning_rate": 4.4456251652127945e-05,
+      "loss": 1.2024,
+      "step": 20982
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8480175733566284,
+      "learning_rate": 4.445598731165742e-05,
+      "loss": 1.2569,
+      "step": 20983
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.8722689151763916,
+      "learning_rate": 4.445572297118689e-05,
+      "loss": 1.3767,
+      "step": 20984
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.8433213233947754,
+      "learning_rate": 4.4455458630716365e-05,
+      "loss": 1.3948,
+      "step": 20985
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2872345447540283,
+      "learning_rate": 4.445519429024584e-05,
+      "loss": 1.2924,
+      "step": 20986
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8001734018325806,
+      "learning_rate": 4.445492994977531e-05,
+      "loss": 1.138,
+      "step": 20987
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7339051961898804,
+      "learning_rate": 4.4454665609304786e-05,
+      "loss": 1.3197,
+      "step": 20988
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.407686471939087,
+      "learning_rate": 4.445440126883426e-05,
+      "loss": 1.2702,
+      "step": 20989
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9702106714248657,
+      "learning_rate": 4.445413692836373e-05,
+      "loss": 1.4959,
+      "step": 20990
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0577497482299805,
+      "learning_rate": 4.445387258789321e-05,
+      "loss": 1.1145,
+      "step": 20991
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.310023546218872,
+      "learning_rate": 4.445360824742269e-05,
+      "loss": 1.3738,
+      "step": 20992
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1932122707366943,
+      "learning_rate": 4.4453343906952154e-05,
+      "loss": 1.234,
+      "step": 20993
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8216139078140259,
+      "learning_rate": 4.4453079566481634e-05,
+      "loss": 1.2543,
+      "step": 20994
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0168824195861816,
+      "learning_rate": 4.44528152260111e-05,
+      "loss": 1.3196,
+      "step": 20995
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.264200210571289,
+      "learning_rate": 4.445255088554058e-05,
+      "loss": 1.5482,
+      "step": 20996
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.415726900100708,
+      "learning_rate": 4.4452286545070055e-05,
+      "loss": 1.3221,
+      "step": 20997
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9626078605651855,
+      "learning_rate": 4.445202220459952e-05,
+      "loss": 1.3916,
+      "step": 20998
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.482296943664551,
+      "learning_rate": 4.4451757864129e-05,
+      "loss": 1.5145,
+      "step": 20999
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7336828708648682,
+      "learning_rate": 4.445149352365847e-05,
+      "loss": 1.1931,
+      "step": 21000
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.155914783477783,
+      "learning_rate": 4.445122918318795e-05,
+      "loss": 1.2734,
+      "step": 21001
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3521878719329834,
+      "learning_rate": 4.445096484271742e-05,
+      "loss": 1.2118,
+      "step": 21002
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.5910931825637817,
+      "learning_rate": 4.4450700502246896e-05,
+      "loss": 1.2364,
+      "step": 21003
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7097105979919434,
+      "learning_rate": 4.445043616177637e-05,
+      "loss": 1.2584,
+      "step": 21004
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3178634643554688,
+      "learning_rate": 4.4450171821305844e-05,
+      "loss": 1.2377,
+      "step": 21005
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0807130336761475,
+      "learning_rate": 4.444990748083532e-05,
+      "loss": 1.4065,
+      "step": 21006
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.462984323501587,
+      "learning_rate": 4.444964314036479e-05,
+      "loss": 1.3978,
+      "step": 21007
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1104273796081543,
+      "learning_rate": 4.4449378799894264e-05,
+      "loss": 1.2156,
+      "step": 21008
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.924619197845459,
+      "learning_rate": 4.444911445942374e-05,
+      "loss": 1.1764,
+      "step": 21009
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9695707559585571,
+      "learning_rate": 4.444885011895321e-05,
+      "loss": 1.2314,
+      "step": 21010
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0090911388397217,
+      "learning_rate": 4.4448585778482685e-05,
+      "loss": 1.4594,
+      "step": 21011
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.124640464782715,
+      "learning_rate": 4.4448321438012165e-05,
+      "loss": 1.3702,
+      "step": 21012
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.4198639392852783,
+      "learning_rate": 4.444805709754163e-05,
+      "loss": 1.3845,
+      "step": 21013
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9802274703979492,
+      "learning_rate": 4.444779275707111e-05,
+      "loss": 1.2732,
+      "step": 21014
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.117675304412842,
+      "learning_rate": 4.444752841660058e-05,
+      "loss": 1.3807,
+      "step": 21015
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.332881450653076,
+      "learning_rate": 4.444726407613006e-05,
+      "loss": 1.315,
+      "step": 21016
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.046736717224121,
+      "learning_rate": 4.444699973565953e-05,
+      "loss": 1.4278,
+      "step": 21017
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.6235628128051758,
+      "learning_rate": 4.444673539518901e-05,
+      "loss": 1.2126,
+      "step": 21018
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.5103769302368164,
+      "learning_rate": 4.444647105471848e-05,
+      "loss": 1.482,
+      "step": 21019
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.6045197248458862,
+      "learning_rate": 4.4446206714247954e-05,
+      "loss": 1.9493,
+      "step": 21020
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7136154174804688,
+      "learning_rate": 4.444594237377743e-05,
+      "loss": 1.328,
+      "step": 21021
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.364147186279297,
+      "learning_rate": 4.44456780333069e-05,
+      "loss": 1.3715,
+      "step": 21022
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.63417387008667,
+      "learning_rate": 4.4445413692836375e-05,
+      "loss": 1.2327,
+      "step": 21023
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3385860919952393,
+      "learning_rate": 4.444514935236585e-05,
+      "loss": 1.3967,
+      "step": 21024
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.969548225402832,
+      "learning_rate": 4.444488501189532e-05,
+      "loss": 1.3299,
+      "step": 21025
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.329730987548828,
+      "learning_rate": 4.4444620671424795e-05,
+      "loss": 1.2917,
+      "step": 21026
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.387040376663208,
+      "learning_rate": 4.4444356330954276e-05,
+      "loss": 1.1932,
+      "step": 21027
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.4656379222869873,
+      "learning_rate": 4.444409199048374e-05,
+      "loss": 1.3096,
+      "step": 21028
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0049197673797607,
+      "learning_rate": 4.444382765001322e-05,
+      "loss": 1.361,
+      "step": 21029
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.7449612617492676,
+      "learning_rate": 4.444356330954269e-05,
+      "loss": 2.3586,
+      "step": 21030
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8985971212387085,
+      "learning_rate": 4.444329896907217e-05,
+      "loss": 1.316,
+      "step": 21031
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9997586011886597,
+      "learning_rate": 4.4443034628601643e-05,
+      "loss": 1.359,
+      "step": 21032
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2864577770233154,
+      "learning_rate": 4.444277028813111e-05,
+      "loss": 1.3902,
+      "step": 21033
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0894722938537598,
+      "learning_rate": 4.444250594766059e-05,
+      "loss": 1.341,
+      "step": 21034
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.510496139526367,
+      "learning_rate": 4.444224160719006e-05,
+      "loss": 1.4538,
+      "step": 21035
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.585112452507019,
+      "learning_rate": 4.444197726671954e-05,
+      "loss": 1.2164,
+      "step": 21036
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.921541452407837,
+      "learning_rate": 4.444171292624901e-05,
+      "loss": 1.3199,
+      "step": 21037
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.566681385040283,
+      "learning_rate": 4.4441448585778485e-05,
+      "loss": 1.1899,
+      "step": 21038
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3919246196746826,
+      "learning_rate": 4.444118424530796e-05,
+      "loss": 1.5786,
+      "step": 21039
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.468778371810913,
+      "learning_rate": 4.444091990483743e-05,
+      "loss": 1.301,
+      "step": 21040
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.155900239944458,
+      "learning_rate": 4.4440655564366906e-05,
+      "loss": 1.1885,
+      "step": 21041
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7834392786026,
+      "learning_rate": 4.4440391223896386e-05,
+      "loss": 1.2644,
+      "step": 21042
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.5820937156677246,
+      "learning_rate": 4.444012688342585e-05,
+      "loss": 1.2891,
+      "step": 21043
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.073549509048462,
+      "learning_rate": 4.4439862542955326e-05,
+      "loss": 1.3048,
+      "step": 21044
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.235398769378662,
+      "learning_rate": 4.44395982024848e-05,
+      "loss": 1.189,
+      "step": 21045
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2476587295532227,
+      "learning_rate": 4.4439333862014273e-05,
+      "loss": 1.2979,
+      "step": 21046
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.905876874923706,
+      "learning_rate": 4.4439069521543754e-05,
+      "loss": 1.2797,
+      "step": 21047
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.365644693374634,
+      "learning_rate": 4.443880518107322e-05,
+      "loss": 1.4374,
+      "step": 21048
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2108776569366455,
+      "learning_rate": 4.44385408406027e-05,
+      "loss": 1.3562,
+      "step": 21049
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.841099739074707,
+      "learning_rate": 4.443827650013217e-05,
+      "loss": 1.177,
+      "step": 21050
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2955899238586426,
+      "learning_rate": 4.443801215966165e-05,
+      "loss": 1.2712,
+      "step": 21051
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1741952896118164,
+      "learning_rate": 4.443774781919112e-05,
+      "loss": 1.2127,
+      "step": 21052
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8290961980819702,
+      "learning_rate": 4.4437483478720595e-05,
+      "loss": 2.0036,
+      "step": 21053
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.169954538345337,
+      "learning_rate": 4.443721913825007e-05,
+      "loss": 1.3789,
+      "step": 21054
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.223094940185547,
+      "learning_rate": 4.443695479777954e-05,
+      "loss": 1.4021,
+      "step": 21055
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.113466739654541,
+      "learning_rate": 4.4436690457309016e-05,
+      "loss": 1.252,
+      "step": 21056
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9350309371948242,
+      "learning_rate": 4.443642611683849e-05,
+      "loss": 1.3491,
+      "step": 21057
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0472278594970703,
+      "learning_rate": 4.443616177636796e-05,
+      "loss": 1.3306,
+      "step": 21058
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.928776741027832,
+      "learning_rate": 4.443589743589744e-05,
+      "loss": 1.5157,
+      "step": 21059
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8597795963287354,
+      "learning_rate": 4.443563309542691e-05,
+      "loss": 1.2542,
+      "step": 21060
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.993927001953125,
+      "learning_rate": 4.4435368754956384e-05,
+      "loss": 1.3112,
+      "step": 21061
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8453387022018433,
+      "learning_rate": 4.4435104414485864e-05,
+      "loss": 1.2576,
+      "step": 21062
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.533644437789917,
+      "learning_rate": 4.443484007401533e-05,
+      "loss": 1.3277,
+      "step": 21063
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8506865501403809,
+      "learning_rate": 4.443457573354481e-05,
+      "loss": 1.1965,
+      "step": 21064
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.001422643661499,
+      "learning_rate": 4.443431139307428e-05,
+      "loss": 1.2184,
+      "step": 21065
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.6366171836853027,
+      "learning_rate": 4.443404705260376e-05,
+      "loss": 1.3827,
+      "step": 21066
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1082849502563477,
+      "learning_rate": 4.443378271213323e-05,
+      "loss": 1.2349,
+      "step": 21067
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.0899925231933594,
+      "learning_rate": 4.4433518371662706e-05,
+      "loss": 1.5389,
+      "step": 21068
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.5851809978485107,
+      "learning_rate": 4.443325403119218e-05,
+      "loss": 1.5052,
+      "step": 21069
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2898731231689453,
+      "learning_rate": 4.4432989690721646e-05,
+      "loss": 1.3265,
+      "step": 21070
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.922235369682312,
+      "learning_rate": 4.4432725350251126e-05,
+      "loss": 1.3146,
+      "step": 21071
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.543797731399536,
+      "learning_rate": 4.44324610097806e-05,
+      "loss": 1.0478,
+      "step": 21072
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.6583611965179443,
+      "learning_rate": 4.4432196669310073e-05,
+      "loss": 1.2687,
+      "step": 21073
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.194599151611328,
+      "learning_rate": 4.443193232883955e-05,
+      "loss": 1.3461,
+      "step": 21074
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.8850693702697754,
+      "learning_rate": 4.443166798836902e-05,
+      "loss": 1.4916,
+      "step": 21075
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.023405075073242,
+      "learning_rate": 4.4431403647898494e-05,
+      "loss": 1.2328,
+      "step": 21076
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.6526896953582764,
+      "learning_rate": 4.4431139307427974e-05,
+      "loss": 1.1845,
+      "step": 21077
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1161630153656006,
+      "learning_rate": 4.443087496695744e-05,
+      "loss": 1.4655,
+      "step": 21078
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7514203786849976,
+      "learning_rate": 4.4430610626486915e-05,
+      "loss": 1.2133,
+      "step": 21079
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7965986728668213,
+      "learning_rate": 4.443034628601639e-05,
+      "loss": 1.2655,
+      "step": 21080
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2043869495391846,
+      "learning_rate": 4.443008194554586e-05,
+      "loss": 1.3989,
+      "step": 21081
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.4103105068206787,
+      "learning_rate": 4.442981760507534e-05,
+      "loss": 1.1713,
+      "step": 21082
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.047497034072876,
+      "learning_rate": 4.442955326460481e-05,
+      "loss": 1.2944,
+      "step": 21083
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.749917984008789,
+      "learning_rate": 4.442928892413429e-05,
+      "loss": 1.3702,
+      "step": 21084
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9302045106887817,
+      "learning_rate": 4.4429024583663756e-05,
+      "loss": 1.2483,
+      "step": 21085
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.9585115909576416,
+      "learning_rate": 4.442876024319324e-05,
+      "loss": 1.2825,
+      "step": 21086
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.70493483543396,
+      "learning_rate": 4.442849590272271e-05,
+      "loss": 1.1534,
+      "step": 21087
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.957224726676941,
+      "learning_rate": 4.4428231562252184e-05,
+      "loss": 1.2867,
+      "step": 21088
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8975415229797363,
+      "learning_rate": 4.442796722178166e-05,
+      "loss": 1.345,
+      "step": 21089
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8209600448608398,
+      "learning_rate": 4.442770288131113e-05,
+      "loss": 1.2803,
+      "step": 21090
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3905632495880127,
+      "learning_rate": 4.4427438540840604e-05,
+      "loss": 1.3191,
+      "step": 21091
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.553053617477417,
+      "learning_rate": 4.442717420037008e-05,
+      "loss": 1.5296,
+      "step": 21092
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.492018222808838,
+      "learning_rate": 4.442690985989955e-05,
+      "loss": 1.3232,
+      "step": 21093
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.8353750705718994,
+      "learning_rate": 4.4426645519429025e-05,
+      "loss": 1.3662,
+      "step": 21094
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2284557819366455,
+      "learning_rate": 4.44263811789585e-05,
+      "loss": 1.1328,
+      "step": 21095
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9082318544387817,
+      "learning_rate": 4.442611683848797e-05,
+      "loss": 1.2493,
+      "step": 21096
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.4665117263793945,
+      "learning_rate": 4.442585249801745e-05,
+      "loss": 1.2514,
+      "step": 21097
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2003605365753174,
+      "learning_rate": 4.442558815754692e-05,
+      "loss": 1.1414,
+      "step": 21098
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.7470386028289795,
+      "learning_rate": 4.44253238170764e-05,
+      "loss": 1.3086,
+      "step": 21099
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.289860725402832,
+      "learning_rate": 4.442505947660587e-05,
+      "loss": 1.2037,
+      "step": 21100
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9950039386749268,
+      "learning_rate": 4.442479513613535e-05,
+      "loss": 1.2452,
+      "step": 21101
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.729148030281067,
+      "learning_rate": 4.442453079566482e-05,
+      "loss": 1.3434,
+      "step": 21102
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.354807138442993,
+      "learning_rate": 4.4424266455194294e-05,
+      "loss": 1.4315,
+      "step": 21103
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0701189041137695,
+      "learning_rate": 4.442400211472377e-05,
+      "loss": 1.3434,
+      "step": 21104
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.938133955001831,
+      "learning_rate": 4.442373777425324e-05,
+      "loss": 1.2473,
+      "step": 21105
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0838794708251953,
+      "learning_rate": 4.4423473433782715e-05,
+      "loss": 1.3376,
+      "step": 21106
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0056517124176025,
+      "learning_rate": 4.442320909331219e-05,
+      "loss": 1.3441,
+      "step": 21107
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3172717094421387,
+      "learning_rate": 4.442294475284166e-05,
+      "loss": 1.2878,
+      "step": 21108
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3058245182037354,
+      "learning_rate": 4.4422680412371136e-05,
+      "loss": 1.2794,
+      "step": 21109
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.4035284519195557,
+      "learning_rate": 4.442241607190061e-05,
+      "loss": 1.499,
+      "step": 21110
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2577316761016846,
+      "learning_rate": 4.442215173143008e-05,
+      "loss": 1.3083,
+      "step": 21111
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9069889783859253,
+      "learning_rate": 4.442188739095956e-05,
+      "loss": 1.7099,
+      "step": 21112
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9167327880859375,
+      "learning_rate": 4.442162305048903e-05,
+      "loss": 1.2772,
+      "step": 21113
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2029192447662354,
+      "learning_rate": 4.442135871001851e-05,
+      "loss": 1.2069,
+      "step": 21114
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9995105266571045,
+      "learning_rate": 4.442109436954798e-05,
+      "loss": 1.2106,
+      "step": 21115
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.17081618309021,
+      "learning_rate": 4.442083002907745e-05,
+      "loss": 1.4233,
+      "step": 21116
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8634469509124756,
+      "learning_rate": 4.442056568860693e-05,
+      "loss": 1.3469,
+      "step": 21117
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8729603290557861,
+      "learning_rate": 4.44203013481364e-05,
+      "loss": 1.2048,
+      "step": 21118
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.710189700126648,
+      "learning_rate": 4.442003700766588e-05,
+      "loss": 1.2751,
+      "step": 21119
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.3413379192352295,
+      "learning_rate": 4.4419772667195345e-05,
+      "loss": 1.5495,
+      "step": 21120
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2958779335021973,
+      "learning_rate": 4.4419508326724825e-05,
+      "loss": 1.2811,
+      "step": 21121
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3158109188079834,
+      "learning_rate": 4.44192439862543e-05,
+      "loss": 1.4237,
+      "step": 21122
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.992261290550232,
+      "learning_rate": 4.441897964578377e-05,
+      "loss": 1.228,
+      "step": 21123
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9501550197601318,
+      "learning_rate": 4.4418715305313246e-05,
+      "loss": 1.281,
+      "step": 21124
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.830232858657837,
+      "learning_rate": 4.441845096484272e-05,
+      "loss": 1.766,
+      "step": 21125
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0624048709869385,
+      "learning_rate": 4.441818662437219e-05,
+      "loss": 1.3478,
+      "step": 21126
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9512733221054077,
+      "learning_rate": 4.4417922283901667e-05,
+      "loss": 1.3098,
+      "step": 21127
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.055131435394287,
+      "learning_rate": 4.441765794343114e-05,
+      "loss": 1.3585,
+      "step": 21128
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.4156978130340576,
+      "learning_rate": 4.4417393602960614e-05,
+      "loss": 1.2604,
+      "step": 21129
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.265554904937744,
+      "learning_rate": 4.441712926249009e-05,
+      "loss": 1.694,
+      "step": 21130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3081068992614746,
+      "learning_rate": 4.441686492201956e-05,
+      "loss": 1.3579,
+      "step": 21131
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7230088710784912,
+      "learning_rate": 4.441660058154904e-05,
+      "loss": 1.2739,
+      "step": 21132
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.782655715942383,
+      "learning_rate": 4.441633624107851e-05,
+      "loss": 1.4432,
+      "step": 21133
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.721792221069336,
+      "learning_rate": 4.441607190060799e-05,
+      "loss": 1.4592,
+      "step": 21134
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8430001735687256,
+      "learning_rate": 4.4415807560137455e-05,
+      "loss": 1.2484,
+      "step": 21135
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8936015367507935,
+      "learning_rate": 4.4415543219666936e-05,
+      "loss": 1.335,
+      "step": 21136
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.316596508026123,
+      "learning_rate": 4.441527887919641e-05,
+      "loss": 1.2898,
+      "step": 21137
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.797872304916382,
+      "learning_rate": 4.441501453872588e-05,
+      "loss": 1.372,
+      "step": 21138
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.024085283279419,
+      "learning_rate": 4.4414750198255356e-05,
+      "loss": 1.2142,
+      "step": 21139
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.349771499633789,
+      "learning_rate": 4.441448585778483e-05,
+      "loss": 1.4069,
+      "step": 21140
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8155573606491089,
+      "learning_rate": 4.44142215173143e-05,
+      "loss": 1.492,
+      "step": 21141
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.90347158908844,
+      "learning_rate": 4.441395717684378e-05,
+      "loss": 1.3837,
+      "step": 21142
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9551420211791992,
+      "learning_rate": 4.441369283637325e-05,
+      "loss": 1.1836,
+      "step": 21143
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.986136794090271,
+      "learning_rate": 4.4413428495902724e-05,
+      "loss": 1.2794,
+      "step": 21144
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.023216962814331,
+      "learning_rate": 4.44131641554322e-05,
+      "loss": 1.1247,
+      "step": 21145
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7693520784378052,
+      "learning_rate": 4.441289981496167e-05,
+      "loss": 1.2349,
+      "step": 21146
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5310959815979004,
+      "learning_rate": 4.441263547449115e-05,
+      "loss": 1.1584,
+      "step": 21147
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1702797412872314,
+      "learning_rate": 4.441237113402062e-05,
+      "loss": 1.3217,
+      "step": 21148
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2892775535583496,
+      "learning_rate": 4.44121067935501e-05,
+      "loss": 1.2871,
+      "step": 21149
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0312399864196777,
+      "learning_rate": 4.4411842453079565e-05,
+      "loss": 1.3955,
+      "step": 21150
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3916521072387695,
+      "learning_rate": 4.441157811260904e-05,
+      "loss": 1.2169,
+      "step": 21151
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0383846759796143,
+      "learning_rate": 4.441131377213852e-05,
+      "loss": 1.3627,
+      "step": 21152
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9844828844070435,
+      "learning_rate": 4.4411049431667986e-05,
+      "loss": 1.25,
+      "step": 21153
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.892445683479309,
+      "learning_rate": 4.4410785091197467e-05,
+      "loss": 1.3133,
+      "step": 21154
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.915785312652588,
+      "learning_rate": 4.441052075072693e-05,
+      "loss": 1.1323,
+      "step": 21155
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7893555164337158,
+      "learning_rate": 4.4410256410256414e-05,
+      "loss": 1.2485,
+      "step": 21156
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5789402723312378,
+      "learning_rate": 4.440999206978589e-05,
+      "loss": 1.1625,
+      "step": 21157
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.020514965057373,
+      "learning_rate": 4.440972772931536e-05,
+      "loss": 1.4046,
+      "step": 21158
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7508314847946167,
+      "learning_rate": 4.4409463388844834e-05,
+      "loss": 1.2041,
+      "step": 21159
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3182427883148193,
+      "learning_rate": 4.440919904837431e-05,
+      "loss": 1.3288,
+      "step": 21160
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9795186519622803,
+      "learning_rate": 4.440893470790378e-05,
+      "loss": 1.3902,
+      "step": 21161
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9418257474899292,
+      "learning_rate": 4.4408670367433255e-05,
+      "loss": 1.1876,
+      "step": 21162
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0673635005950928,
+      "learning_rate": 4.440840602696273e-05,
+      "loss": 1.3796,
+      "step": 21163
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.080005407333374,
+      "learning_rate": 4.44081416864922e-05,
+      "loss": 1.257,
+      "step": 21164
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7937756776809692,
+      "learning_rate": 4.4407877346021676e-05,
+      "loss": 1.3003,
+      "step": 21165
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2463812828063965,
+      "learning_rate": 4.440761300555115e-05,
+      "loss": 1.3753,
+      "step": 21166
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9308584928512573,
+      "learning_rate": 4.440734866508063e-05,
+      "loss": 1.2214,
+      "step": 21167
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.6048355102539062,
+      "learning_rate": 4.4407084324610097e-05,
+      "loss": 1.1614,
+      "step": 21168
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5225924253463745,
+      "learning_rate": 4.440681998413958e-05,
+      "loss": 1.1247,
+      "step": 21169
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3049306869506836,
+      "learning_rate": 4.4406555643669044e-05,
+      "loss": 1.3085,
+      "step": 21170
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9675086736679077,
+      "learning_rate": 4.4406291303198524e-05,
+      "loss": 1.1753,
+      "step": 21171
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1370129585266113,
+      "learning_rate": 4.4406026962728e-05,
+      "loss": 1.2899,
+      "step": 21172
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.6830527782440186,
+      "learning_rate": 4.440576262225747e-05,
+      "loss": 1.3476,
+      "step": 21173
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5963315963745117,
+      "learning_rate": 4.4405498281786945e-05,
+      "loss": 1.2073,
+      "step": 21174
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.599186658859253,
+      "learning_rate": 4.440523394131642e-05,
+      "loss": 1.442,
+      "step": 21175
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7670419216156006,
+      "learning_rate": 4.440496960084589e-05,
+      "loss": 1.2939,
+      "step": 21176
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.338327407836914,
+      "learning_rate": 4.4404705260375365e-05,
+      "loss": 1.4005,
+      "step": 21177
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5180925130844116,
+      "learning_rate": 4.440444091990484e-05,
+      "loss": 1.2131,
+      "step": 21178
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.4795851707458496,
+      "learning_rate": 4.440417657943431e-05,
+      "loss": 1.6701,
+      "step": 21179
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9295575618743896,
+      "learning_rate": 4.4403912238963786e-05,
+      "loss": 1.2892,
+      "step": 21180
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8705956935882568,
+      "learning_rate": 4.440364789849326e-05,
+      "loss": 1.3292,
+      "step": 21181
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1370913982391357,
+      "learning_rate": 4.440338355802274e-05,
+      "loss": 1.2074,
+      "step": 21182
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8950161933898926,
+      "learning_rate": 4.440311921755221e-05,
+      "loss": 1.2266,
+      "step": 21183
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1185948848724365,
+      "learning_rate": 4.440285487708169e-05,
+      "loss": 1.2601,
+      "step": 21184
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0217483043670654,
+      "learning_rate": 4.4402590536611154e-05,
+      "loss": 1.3022,
+      "step": 21185
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0954267978668213,
+      "learning_rate": 4.4402326196140634e-05,
+      "loss": 1.3715,
+      "step": 21186
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7890033721923828,
+      "learning_rate": 4.440206185567011e-05,
+      "loss": 1.3512,
+      "step": 21187
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.6039502620697021,
+      "learning_rate": 4.4401797515199575e-05,
+      "loss": 1.1516,
+      "step": 21188
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.472884178161621,
+      "learning_rate": 4.4401533174729055e-05,
+      "loss": 1.4808,
+      "step": 21189
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8078386783599854,
+      "learning_rate": 4.440126883425852e-05,
+      "loss": 1.2346,
+      "step": 21190
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3628780841827393,
+      "learning_rate": 4.4401004493788e-05,
+      "loss": 1.3021,
+      "step": 21191
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7081032991409302,
+      "learning_rate": 4.4400740153317476e-05,
+      "loss": 1.2074,
+      "step": 21192
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.085003137588501,
+      "learning_rate": 4.440047581284695e-05,
+      "loss": 1.1068,
+      "step": 21193
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.090674638748169,
+      "learning_rate": 4.440021147237642e-05,
+      "loss": 1.4105,
+      "step": 21194
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.810615301132202,
+      "learning_rate": 4.4399947131905897e-05,
+      "loss": 1.4117,
+      "step": 21195
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3507015705108643,
+      "learning_rate": 4.439968279143537e-05,
+      "loss": 1.3403,
+      "step": 21196
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0748207569122314,
+      "learning_rate": 4.4399418450964844e-05,
+      "loss": 1.1928,
+      "step": 21197
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9606132507324219,
+      "learning_rate": 4.439915411049432e-05,
+      "loss": 1.3544,
+      "step": 21198
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.799635887145996,
+      "learning_rate": 4.439888977002379e-05,
+      "loss": 1.3571,
+      "step": 21199
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3776910305023193,
+      "learning_rate": 4.4398625429553264e-05,
+      "loss": 1.4304,
+      "step": 21200
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9698588848114014,
+      "learning_rate": 4.439836108908274e-05,
+      "loss": 1.2955,
+      "step": 21201
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.4684040546417236,
+      "learning_rate": 4.439809674861222e-05,
+      "loss": 1.3856,
+      "step": 21202
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.921167254447937,
+      "learning_rate": 4.4397832408141685e-05,
+      "loss": 1.2991,
+      "step": 21203
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8801850080490112,
+      "learning_rate": 4.4397568067671165e-05,
+      "loss": 1.2266,
+      "step": 21204
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.146411657333374,
+      "learning_rate": 4.439730372720063e-05,
+      "loss": 1.2357,
+      "step": 21205
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.45040225982666,
+      "learning_rate": 4.439703938673011e-05,
+      "loss": 1.4086,
+      "step": 21206
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.87858247756958,
+      "learning_rate": 4.4396775046259586e-05,
+      "loss": 1.4458,
+      "step": 21207
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9286938905715942,
+      "learning_rate": 4.439651070578906e-05,
+      "loss": 1.2272,
+      "step": 21208
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2508745193481445,
+      "learning_rate": 4.439624636531853e-05,
+      "loss": 1.4181,
+      "step": 21209
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0997025966644287,
+      "learning_rate": 4.439598202484801e-05,
+      "loss": 1.3769,
+      "step": 21210
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7535765171051025,
+      "learning_rate": 4.439571768437748e-05,
+      "loss": 1.0517,
+      "step": 21211
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8694382905960083,
+      "learning_rate": 4.4395453343906954e-05,
+      "loss": 1.3081,
+      "step": 21212
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.598795771598816,
+      "learning_rate": 4.439518900343643e-05,
+      "loss": 1.1922,
+      "step": 21213
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.888808250427246,
+      "learning_rate": 4.43949246629659e-05,
+      "loss": 1.1756,
+      "step": 21214
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0707883834838867,
+      "learning_rate": 4.4394660322495375e-05,
+      "loss": 1.1553,
+      "step": 21215
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.6889842748641968,
+      "learning_rate": 4.439439598202485e-05,
+      "loss": 1.2474,
+      "step": 21216
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7082997560501099,
+      "learning_rate": 4.439413164155433e-05,
+      "loss": 2.277,
+      "step": 21217
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.083108901977539,
+      "learning_rate": 4.4393867301083795e-05,
+      "loss": 1.358,
+      "step": 21218
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1206164360046387,
+      "learning_rate": 4.4393602960613276e-05,
+      "loss": 1.197,
+      "step": 21219
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.888522744178772,
+      "learning_rate": 4.439333862014274e-05,
+      "loss": 1.3543,
+      "step": 21220
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8490482568740845,
+      "learning_rate": 4.439307427967222e-05,
+      "loss": 1.2539,
+      "step": 21221
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.6977131366729736,
+      "learning_rate": 4.4392809939201696e-05,
+      "loss": 1.3141,
+      "step": 21222
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.258100986480713,
+      "learning_rate": 4.439254559873117e-05,
+      "loss": 1.3417,
+      "step": 21223
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0087456703186035,
+      "learning_rate": 4.4392281258260644e-05,
+      "loss": 1.2984,
+      "step": 21224
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5548208951950073,
+      "learning_rate": 4.439201691779011e-05,
+      "loss": 1.8441,
+      "step": 21225
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.209725856781006,
+      "learning_rate": 4.439175257731959e-05,
+      "loss": 1.1812,
+      "step": 21226
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.6675571203231812,
+      "learning_rate": 4.4391488236849064e-05,
+      "loss": 1.1834,
+      "step": 21227
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.799393653869629,
+      "learning_rate": 4.439122389637854e-05,
+      "loss": 1.7443,
+      "step": 21228
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9495340585708618,
+      "learning_rate": 4.439095955590801e-05,
+      "loss": 1.2991,
+      "step": 21229
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1488406658172607,
+      "learning_rate": 4.4390695215437485e-05,
+      "loss": 1.3039,
+      "step": 21230
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.851844072341919,
+      "learning_rate": 4.439043087496696e-05,
+      "loss": 1.2171,
+      "step": 21231
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.8306899070739746,
+      "learning_rate": 4.439016653449644e-05,
+      "loss": 1.4742,
+      "step": 21232
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.059372901916504,
+      "learning_rate": 4.4389902194025906e-05,
+      "loss": 1.2201,
+      "step": 21233
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1274361610412598,
+      "learning_rate": 4.438963785355538e-05,
+      "loss": 1.3526,
+      "step": 21234
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.167062997817993,
+      "learning_rate": 4.438937351308485e-05,
+      "loss": 1.3375,
+      "step": 21235
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5640333890914917,
+      "learning_rate": 4.4389109172614326e-05,
+      "loss": 1.207,
+      "step": 21236
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.083845615386963,
+      "learning_rate": 4.438884483214381e-05,
+      "loss": 1.3501,
+      "step": 21237
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3085477352142334,
+      "learning_rate": 4.4388580491673274e-05,
+      "loss": 1.4852,
+      "step": 21238
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2497198581695557,
+      "learning_rate": 4.4388316151202754e-05,
+      "loss": 1.2167,
+      "step": 21239
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8604069948196411,
+      "learning_rate": 4.438805181073222e-05,
+      "loss": 1.3962,
+      "step": 21240
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8577451705932617,
+      "learning_rate": 4.43877874702617e-05,
+      "loss": 1.2153,
+      "step": 21241
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5438792705535889,
+      "learning_rate": 4.4387523129791175e-05,
+      "loss": 1.1592,
+      "step": 21242
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.798895835876465,
+      "learning_rate": 4.438725878932065e-05,
+      "loss": 1.3531,
+      "step": 21243
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3997106552124023,
+      "learning_rate": 4.438699444885012e-05,
+      "loss": 1.227,
+      "step": 21244
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5518262386322021,
+      "learning_rate": 4.4386730108379595e-05,
+      "loss": 1.3085,
+      "step": 21245
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9250359535217285,
+      "learning_rate": 4.438646576790907e-05,
+      "loss": 1.2835,
+      "step": 21246
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9722338914871216,
+      "learning_rate": 4.438620142743854e-05,
+      "loss": 1.4104,
+      "step": 21247
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7959436178207397,
+      "learning_rate": 4.4385937086968016e-05,
+      "loss": 1.2203,
+      "step": 21248
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.7626919746398926,
+      "learning_rate": 4.438567274649749e-05,
+      "loss": 1.2094,
+      "step": 21249
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9647682905197144,
+      "learning_rate": 4.438540840602696e-05,
+      "loss": 1.3479,
+      "step": 21250
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8793154954910278,
+      "learning_rate": 4.438514406555644e-05,
+      "loss": 1.3058,
+      "step": 21251
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9499433040618896,
+      "learning_rate": 4.438487972508592e-05,
+      "loss": 2.1997,
+      "step": 21252
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.6363131999969482,
+      "learning_rate": 4.4384615384615384e-05,
+      "loss": 1.8615,
+      "step": 21253
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0379960536956787,
+      "learning_rate": 4.4384351044144864e-05,
+      "loss": 1.2043,
+      "step": 21254
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.6270873546600342,
+      "learning_rate": 4.438408670367433e-05,
+      "loss": 1.2547,
+      "step": 21255
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9406180381774902,
+      "learning_rate": 4.438382236320381e-05,
+      "loss": 1.3358,
+      "step": 21256
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.111762285232544,
+      "learning_rate": 4.4383558022733285e-05,
+      "loss": 1.3048,
+      "step": 21257
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8801158666610718,
+      "learning_rate": 4.438329368226276e-05,
+      "loss": 1.4182,
+      "step": 21258
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0409281253814697,
+      "learning_rate": 4.438302934179223e-05,
+      "loss": 1.2815,
+      "step": 21259
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.142190456390381,
+      "learning_rate": 4.43827650013217e-05,
+      "loss": 1.3038,
+      "step": 21260
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.101173162460327,
+      "learning_rate": 4.438250066085118e-05,
+      "loss": 1.2431,
+      "step": 21261
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0980336666107178,
+      "learning_rate": 4.438223632038065e-05,
+      "loss": 1.3682,
+      "step": 21262
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.29082989692688,
+      "learning_rate": 4.4381971979910126e-05,
+      "loss": 1.4084,
+      "step": 21263
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.160090208053589,
+      "learning_rate": 4.43817076394396e-05,
+      "loss": 1.9413,
+      "step": 21264
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7964646816253662,
+      "learning_rate": 4.4381443298969074e-05,
+      "loss": 1.2682,
+      "step": 21265
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1972248554229736,
+      "learning_rate": 4.438117895849855e-05,
+      "loss": 1.4205,
+      "step": 21266
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7155814170837402,
+      "learning_rate": 4.438091461802803e-05,
+      "loss": 1.3626,
+      "step": 21267
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.4486286640167236,
+      "learning_rate": 4.4380650277557494e-05,
+      "loss": 1.2746,
+      "step": 21268
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.133192777633667,
+      "learning_rate": 4.438038593708697e-05,
+      "loss": 1.2707,
+      "step": 21269
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.4511971473693848,
+      "learning_rate": 4.438012159661644e-05,
+      "loss": 1.0844,
+      "step": 21270
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3999524116516113,
+      "learning_rate": 4.4379857256145915e-05,
+      "loss": 2.4209,
+      "step": 21271
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.0525310039520264,
+      "learning_rate": 4.4379592915675395e-05,
+      "loss": 1.2281,
+      "step": 21272
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.2043139934539795,
+      "learning_rate": 4.437932857520486e-05,
+      "loss": 1.5844,
+      "step": 21273
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.198915481567383,
+      "learning_rate": 4.437906423473434e-05,
+      "loss": 1.1357,
+      "step": 21274
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7704211473464966,
+      "learning_rate": 4.437879989426381e-05,
+      "loss": 1.2414,
+      "step": 21275
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.048799753189087,
+      "learning_rate": 4.437853555379329e-05,
+      "loss": 1.3053,
+      "step": 21276
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7058203220367432,
+      "learning_rate": 4.437827121332276e-05,
+      "loss": 1.2252,
+      "step": 21277
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.6922699213027954,
+      "learning_rate": 4.437800687285224e-05,
+      "loss": 1.4123,
+      "step": 21278
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9111764430999756,
+      "learning_rate": 4.437774253238171e-05,
+      "loss": 1.2932,
+      "step": 21279
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.7974042892456055,
+      "learning_rate": 4.4377478191911184e-05,
+      "loss": 1.4697,
+      "step": 21280
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.808606743812561,
+      "learning_rate": 4.437721385144066e-05,
+      "loss": 1.2327,
+      "step": 21281
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9093111753463745,
+      "learning_rate": 4.437694951097013e-05,
+      "loss": 1.4601,
+      "step": 21282
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1080050468444824,
+      "learning_rate": 4.4376685170499605e-05,
+      "loss": 1.2531,
+      "step": 21283
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.002196788787842,
+      "learning_rate": 4.437642083002908e-05,
+      "loss": 1.3414,
+      "step": 21284
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.6795976161956787,
+      "learning_rate": 4.437615648955855e-05,
+      "loss": 1.0386,
+      "step": 21285
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6219472885131836,
+      "learning_rate": 4.4375892149088025e-05,
+      "loss": 1.2287,
+      "step": 21286
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1341335773468018,
+      "learning_rate": 4.4375627808617506e-05,
+      "loss": 1.1519,
+      "step": 21287
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9891102313995361,
+      "learning_rate": 4.437536346814697e-05,
+      "loss": 1.3102,
+      "step": 21288
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7386280298233032,
+      "learning_rate": 4.437509912767645e-05,
+      "loss": 1.1898,
+      "step": 21289
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.370290517807007,
+      "learning_rate": 4.437483478720592e-05,
+      "loss": 1.4368,
+      "step": 21290
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.713663935661316,
+      "learning_rate": 4.43745704467354e-05,
+      "loss": 1.3302,
+      "step": 21291
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.135775327682495,
+      "learning_rate": 4.4374306106264874e-05,
+      "loss": 1.3917,
+      "step": 21292
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1885006427764893,
+      "learning_rate": 4.437404176579435e-05,
+      "loss": 1.3479,
+      "step": 21293
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.278859853744507,
+      "learning_rate": 4.437377742532382e-05,
+      "loss": 1.3314,
+      "step": 21294
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7391035556793213,
+      "learning_rate": 4.4373513084853294e-05,
+      "loss": 1.3484,
+      "step": 21295
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.6372867822647095,
+      "learning_rate": 4.437324874438277e-05,
+      "loss": 1.1691,
+      "step": 21296
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0568106174468994,
+      "learning_rate": 4.437298440391224e-05,
+      "loss": 1.2801,
+      "step": 21297
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.5960830450057983,
+      "learning_rate": 4.4372720063441715e-05,
+      "loss": 1.2319,
+      "step": 21298
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2356960773468018,
+      "learning_rate": 4.437245572297119e-05,
+      "loss": 1.4251,
+      "step": 21299
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.191384792327881,
+      "learning_rate": 4.437219138250066e-05,
+      "loss": 1.4421,
+      "step": 21300
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.3481192588806152,
+      "learning_rate": 4.4371927042030136e-05,
+      "loss": 1.4893,
+      "step": 21301
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.822273850440979,
+      "learning_rate": 4.4371662701559616e-05,
+      "loss": 1.2285,
+      "step": 21302
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9093515872955322,
+      "learning_rate": 4.437139836108908e-05,
+      "loss": 1.2905,
+      "step": 21303
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.545207977294922,
+      "learning_rate": 4.437113402061856e-05,
+      "loss": 1.4616,
+      "step": 21304
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.5829801559448242,
+      "learning_rate": 4.437086968014803e-05,
+      "loss": 1.0823,
+      "step": 21305
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7375047206878662,
+      "learning_rate": 4.4370605339677504e-05,
+      "loss": 1.3189,
+      "step": 21306
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9312396049499512,
+      "learning_rate": 4.4370340999206984e-05,
+      "loss": 1.2053,
+      "step": 21307
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.812727928161621,
+      "learning_rate": 4.437007665873645e-05,
+      "loss": 1.2432,
+      "step": 21308
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9534552097320557,
+      "learning_rate": 4.436981231826593e-05,
+      "loss": 1.2145,
+      "step": 21309
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8755643367767334,
+      "learning_rate": 4.43695479777954e-05,
+      "loss": 1.2764,
+      "step": 21310
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7971789836883545,
+      "learning_rate": 4.436928363732488e-05,
+      "loss": 1.2273,
+      "step": 21311
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.850017786026001,
+      "learning_rate": 4.436901929685435e-05,
+      "loss": 1.1234,
+      "step": 21312
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8521099090576172,
+      "learning_rate": 4.4368754956383825e-05,
+      "loss": 1.2703,
+      "step": 21313
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.6426219940185547,
+      "learning_rate": 4.43684906159133e-05,
+      "loss": 1.1965,
+      "step": 21314
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0344300270080566,
+      "learning_rate": 4.436822627544277e-05,
+      "loss": 1.1838,
+      "step": 21315
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.20231294631958,
+      "learning_rate": 4.4367961934972246e-05,
+      "loss": 1.3358,
+      "step": 21316
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7978925704956055,
+      "learning_rate": 4.436769759450172e-05,
+      "loss": 1.9233,
+      "step": 21317
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.068420886993408,
+      "learning_rate": 4.436743325403119e-05,
+      "loss": 1.7674,
+      "step": 21318
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.6962159872055054,
+      "learning_rate": 4.436716891356067e-05,
+      "loss": 1.1624,
+      "step": 21319
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8702189922332764,
+      "learning_rate": 4.436690457309014e-05,
+      "loss": 2.3677,
+      "step": 21320
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9974497556686401,
+      "learning_rate": 4.4366640232619614e-05,
+      "loss": 1.2835,
+      "step": 21321
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6168580055236816,
+      "learning_rate": 4.4366375892149094e-05,
+      "loss": 1.4155,
+      "step": 21322
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.439652442932129,
+      "learning_rate": 4.436611155167856e-05,
+      "loss": 1.888,
+      "step": 21323
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7325021028518677,
+      "learning_rate": 4.436584721120804e-05,
+      "loss": 1.2809,
+      "step": 21324
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.082334518432617,
+      "learning_rate": 4.436558287073751e-05,
+      "loss": 1.2698,
+      "step": 21325
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.528026819229126,
+      "learning_rate": 4.436531853026699e-05,
+      "loss": 1.4195,
+      "step": 21326
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7607121467590332,
+      "learning_rate": 4.436505418979646e-05,
+      "loss": 1.2059,
+      "step": 21327
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7670894861221313,
+      "learning_rate": 4.4364789849325936e-05,
+      "loss": 1.2525,
+      "step": 21328
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8101379871368408,
+      "learning_rate": 4.436452550885541e-05,
+      "loss": 1.1402,
+      "step": 21329
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.617201805114746,
+      "learning_rate": 4.436426116838488e-05,
+      "loss": 2.1076,
+      "step": 21330
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2786972522735596,
+      "learning_rate": 4.4363996827914356e-05,
+      "loss": 1.3655,
+      "step": 21331
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.640188694000244,
+      "learning_rate": 4.436373248744383e-05,
+      "loss": 1.4034,
+      "step": 21332
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.05322003364563,
+      "learning_rate": 4.4363468146973304e-05,
+      "loss": 1.25,
+      "step": 21333
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.96694815158844,
+      "learning_rate": 4.436320380650278e-05,
+      "loss": 1.3525,
+      "step": 21334
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1050684452056885,
+      "learning_rate": 4.436293946603225e-05,
+      "loss": 1.1573,
+      "step": 21335
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.034630060195923,
+      "learning_rate": 4.4362675125561724e-05,
+      "loss": 1.4009,
+      "step": 21336
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8053436279296875,
+      "learning_rate": 4.4362410785091205e-05,
+      "loss": 1.1785,
+      "step": 21337
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8766098022460938,
+      "learning_rate": 4.436214644462067e-05,
+      "loss": 1.2844,
+      "step": 21338
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9787184000015259,
+      "learning_rate": 4.436188210415015e-05,
+      "loss": 1.2423,
+      "step": 21339
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8234771490097046,
+      "learning_rate": 4.436161776367962e-05,
+      "loss": 1.2946,
+      "step": 21340
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9146686792373657,
+      "learning_rate": 4.43613534232091e-05,
+      "loss": 1.3822,
+      "step": 21341
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.3791568279266357,
+      "learning_rate": 4.436108908273857e-05,
+      "loss": 1.2459,
+      "step": 21342
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1060705184936523,
+      "learning_rate": 4.436082474226804e-05,
+      "loss": 1.263,
+      "step": 21343
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7919293642044067,
+      "learning_rate": 4.436056040179752e-05,
+      "loss": 1.2387,
+      "step": 21344
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7862093448638916,
+      "learning_rate": 4.4360296061326986e-05,
+      "loss": 1.2591,
+      "step": 21345
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7432328462600708,
+      "learning_rate": 4.436003172085647e-05,
+      "loss": 1.2732,
+      "step": 21346
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9525115489959717,
+      "learning_rate": 4.435976738038594e-05,
+      "loss": 1.3179,
+      "step": 21347
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.05692195892334,
+      "learning_rate": 4.4359503039915414e-05,
+      "loss": 1.2748,
+      "step": 21348
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0109033584594727,
+      "learning_rate": 4.435923869944489e-05,
+      "loss": 1.4187,
+      "step": 21349
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.61477792263031,
+      "learning_rate": 4.435897435897436e-05,
+      "loss": 1.1797,
+      "step": 21350
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.287682294845581,
+      "learning_rate": 4.4358710018503835e-05,
+      "loss": 1.2705,
+      "step": 21351
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0053741931915283,
+      "learning_rate": 4.435844567803331e-05,
+      "loss": 1.3834,
+      "step": 21352
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.3648064136505127,
+      "learning_rate": 4.435818133756278e-05,
+      "loss": 1.3522,
+      "step": 21353
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.19477915763855,
+      "learning_rate": 4.4357916997092255e-05,
+      "loss": 1.2252,
+      "step": 21354
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0228848457336426,
+      "learning_rate": 4.435765265662173e-05,
+      "loss": 1.3153,
+      "step": 21355
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.3229429721832275,
+      "learning_rate": 4.43573883161512e-05,
+      "loss": 1.434,
+      "step": 21356
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0745010375976562,
+      "learning_rate": 4.435712397568068e-05,
+      "loss": 1.2847,
+      "step": 21357
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8907034397125244,
+      "learning_rate": 4.435685963521015e-05,
+      "loss": 1.1845,
+      "step": 21358
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2452456951141357,
+      "learning_rate": 4.435659529473963e-05,
+      "loss": 1.3451,
+      "step": 21359
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.4164893627166748,
+      "learning_rate": 4.43563309542691e-05,
+      "loss": 1.0247,
+      "step": 21360
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9617037773132324,
+      "learning_rate": 4.435606661379858e-05,
+      "loss": 1.2199,
+      "step": 21361
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2869677543640137,
+      "learning_rate": 4.435580227332805e-05,
+      "loss": 1.3765,
+      "step": 21362
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.352888584136963,
+      "learning_rate": 4.4355537932857524e-05,
+      "loss": 1.5154,
+      "step": 21363
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1783783435821533,
+      "learning_rate": 4.4355273592387e-05,
+      "loss": 1.3511,
+      "step": 21364
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1732234954833984,
+      "learning_rate": 4.435500925191647e-05,
+      "loss": 1.3947,
+      "step": 21365
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.127021312713623,
+      "learning_rate": 4.4354744911445945e-05,
+      "loss": 1.3639,
+      "step": 21366
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.312767267227173,
+      "learning_rate": 4.435448057097542e-05,
+      "loss": 1.3039,
+      "step": 21367
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7053226232528687,
+      "learning_rate": 4.435421623050489e-05,
+      "loss": 1.29,
+      "step": 21368
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.643546462059021,
+      "learning_rate": 4.4353951890034366e-05,
+      "loss": 1.2234,
+      "step": 21369
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.223285675048828,
+      "learning_rate": 4.435368754956384e-05,
+      "loss": 1.3683,
+      "step": 21370
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.7387008666992188,
+      "learning_rate": 4.435342320909331e-05,
+      "loss": 1.489,
+      "step": 21371
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1922690868377686,
+      "learning_rate": 4.435315886862279e-05,
+      "loss": 1.2883,
+      "step": 21372
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.030474901199341,
+      "learning_rate": 4.435289452815226e-05,
+      "loss": 1.7038,
+      "step": 21373
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.9474611282348633,
+      "learning_rate": 4.435263018768174e-05,
+      "loss": 1.3867,
+      "step": 21374
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.5374526977539062,
+      "learning_rate": 4.435236584721121e-05,
+      "loss": 1.2543,
+      "step": 21375
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.46608567237854,
+      "learning_rate": 4.435210150674069e-05,
+      "loss": 1.3455,
+      "step": 21376
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.6541714668273926,
+      "learning_rate": 4.435183716627016e-05,
+      "loss": 1.1134,
+      "step": 21377
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.649568557739258,
+      "learning_rate": 4.435157282579963e-05,
+      "loss": 1.392,
+      "step": 21378
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1019251346588135,
+      "learning_rate": 4.435130848532911e-05,
+      "loss": 1.652,
+      "step": 21379
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.038970470428467,
+      "learning_rate": 4.4351044144858575e-05,
+      "loss": 1.2855,
+      "step": 21380
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.884832739830017,
+      "learning_rate": 4.4350779804388055e-05,
+      "loss": 1.1371,
+      "step": 21381
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2807741165161133,
+      "learning_rate": 4.435051546391753e-05,
+      "loss": 1.3458,
+      "step": 21382
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.210395574569702,
+      "learning_rate": 4.4350251123447e-05,
+      "loss": 1.3307,
+      "step": 21383
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.923598527908325,
+      "learning_rate": 4.4349986782976476e-05,
+      "loss": 1.2552,
+      "step": 21384
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.229461908340454,
+      "learning_rate": 4.434972244250595e-05,
+      "loss": 1.4034,
+      "step": 21385
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.068761110305786,
+      "learning_rate": 4.434945810203542e-05,
+      "loss": 1.2762,
+      "step": 21386
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.813260555267334,
+      "learning_rate": 4.43491937615649e-05,
+      "loss": 1.2669,
+      "step": 21387
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.056492328643799,
+      "learning_rate": 4.434892942109437e-05,
+      "loss": 1.3785,
+      "step": 21388
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.480790853500366,
+      "learning_rate": 4.4348665080623844e-05,
+      "loss": 1.3827,
+      "step": 21389
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.7245867252349854,
+      "learning_rate": 4.434840074015332e-05,
+      "loss": 1.3317,
+      "step": 21390
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0723869800567627,
+      "learning_rate": 4.434813639968279e-05,
+      "loss": 1.3141,
+      "step": 21391
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7953112125396729,
+      "learning_rate": 4.434787205921227e-05,
+      "loss": 1.3286,
+      "step": 21392
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.124347448348999,
+      "learning_rate": 4.434760771874174e-05,
+      "loss": 1.3327,
+      "step": 21393
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.012169599533081,
+      "learning_rate": 4.434734337827122e-05,
+      "loss": 1.3411,
+      "step": 21394
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.837172508239746,
+      "learning_rate": 4.4347079037800685e-05,
+      "loss": 1.3365,
+      "step": 21395
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9251506328582764,
+      "learning_rate": 4.4346814697330166e-05,
+      "loss": 1.3529,
+      "step": 21396
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0885088443756104,
+      "learning_rate": 4.434655035685964e-05,
+      "loss": 1.293,
+      "step": 21397
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9854861497879028,
+      "learning_rate": 4.434628601638911e-05,
+      "loss": 1.142,
+      "step": 21398
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1500539779663086,
+      "learning_rate": 4.4346021675918586e-05,
+      "loss": 1.5209,
+      "step": 21399
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0496749877929688,
+      "learning_rate": 4.434575733544806e-05,
+      "loss": 1.2583,
+      "step": 21400
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2957563400268555,
+      "learning_rate": 4.4345492994977533e-05,
+      "loss": 1.0844,
+      "step": 21401
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8145045042037964,
+      "learning_rate": 4.434522865450701e-05,
+      "loss": 1.2916,
+      "step": 21402
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.257415533065796,
+      "learning_rate": 4.434496431403648e-05,
+      "loss": 1.1367,
+      "step": 21403
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.173203706741333,
+      "learning_rate": 4.4344699973565954e-05,
+      "loss": 1.2944,
+      "step": 21404
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.786861538887024,
+      "learning_rate": 4.434443563309543e-05,
+      "loss": 1.1307,
+      "step": 21405
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7371728420257568,
+      "learning_rate": 4.43441712926249e-05,
+      "loss": 1.2404,
+      "step": 21406
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2320046424865723,
+      "learning_rate": 4.434390695215438e-05,
+      "loss": 1.3673,
+      "step": 21407
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6781561374664307,
+      "learning_rate": 4.434364261168385e-05,
+      "loss": 1.4498,
+      "step": 21408
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8010517358779907,
+      "learning_rate": 4.434337827121333e-05,
+      "loss": 1.3322,
+      "step": 21409
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.016906261444092,
+      "learning_rate": 4.4343113930742796e-05,
+      "loss": 1.2988,
+      "step": 21410
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.851323127746582,
+      "learning_rate": 4.4342849590272276e-05,
+      "loss": 2.2084,
+      "step": 21411
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9701786041259766,
+      "learning_rate": 4.434258524980175e-05,
+      "loss": 1.2873,
+      "step": 21412
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9891729354858398,
+      "learning_rate": 4.434232090933122e-05,
+      "loss": 1.3071,
+      "step": 21413
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1459758281707764,
+      "learning_rate": 4.43420565688607e-05,
+      "loss": 1.2211,
+      "step": 21414
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1397361755371094,
+      "learning_rate": 4.4341792228390163e-05,
+      "loss": 1.179,
+      "step": 21415
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0977680683135986,
+      "learning_rate": 4.4341527887919644e-05,
+      "loss": 1.1812,
+      "step": 21416
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7085540294647217,
+      "learning_rate": 4.434126354744912e-05,
+      "loss": 1.3466,
+      "step": 21417
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2196829319000244,
+      "learning_rate": 4.434099920697859e-05,
+      "loss": 2.3786,
+      "step": 21418
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7481889724731445,
+      "learning_rate": 4.4340734866508064e-05,
+      "loss": 1.2796,
+      "step": 21419
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1332194805145264,
+      "learning_rate": 4.434047052603754e-05,
+      "loss": 1.1577,
+      "step": 21420
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.1342854499816895,
+      "learning_rate": 4.434020618556701e-05,
+      "loss": 1.5178,
+      "step": 21421
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0595216751098633,
+      "learning_rate": 4.433994184509649e-05,
+      "loss": 1.1988,
+      "step": 21422
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.299542188644409,
+      "learning_rate": 4.433967750462596e-05,
+      "loss": 1.1994,
+      "step": 21423
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.14780330657959,
+      "learning_rate": 4.433941316415543e-05,
+      "loss": 1.3027,
+      "step": 21424
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9230817556381226,
+      "learning_rate": 4.4339148823684906e-05,
+      "loss": 1.4466,
+      "step": 21425
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.035066843032837,
+      "learning_rate": 4.433888448321438e-05,
+      "loss": 1.2837,
+      "step": 21426
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0818140506744385,
+      "learning_rate": 4.433862014274386e-05,
+      "loss": 1.2708,
+      "step": 21427
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7731581926345825,
+      "learning_rate": 4.433835580227333e-05,
+      "loss": 1.5975,
+      "step": 21428
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.9217448234558105,
+      "learning_rate": 4.433809146180281e-05,
+      "loss": 1.4216,
+      "step": 21429
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1878223419189453,
+      "learning_rate": 4.4337827121332274e-05,
+      "loss": 1.3075,
+      "step": 21430
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.163438558578491,
+      "learning_rate": 4.4337562780861754e-05,
+      "loss": 1.2104,
+      "step": 21431
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.809097170829773,
+      "learning_rate": 4.433729844039123e-05,
+      "loss": 1.1596,
+      "step": 21432
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.3792734146118164,
+      "learning_rate": 4.43370340999207e-05,
+      "loss": 1.2137,
+      "step": 21433
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8831075429916382,
+      "learning_rate": 4.4336769759450175e-05,
+      "loss": 1.3448,
+      "step": 21434
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.36179518699646,
+      "learning_rate": 4.433650541897965e-05,
+      "loss": 1.3193,
+      "step": 21435
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.472029447555542,
+      "learning_rate": 4.433624107850912e-05,
+      "loss": 1.3625,
+      "step": 21436
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6769917011260986,
+      "learning_rate": 4.4335976738038596e-05,
+      "loss": 1.6284,
+      "step": 21437
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.938819408416748,
+      "learning_rate": 4.433571239756807e-05,
+      "loss": 1.3253,
+      "step": 21438
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2428157329559326,
+      "learning_rate": 4.433544805709754e-05,
+      "loss": 1.4681,
+      "step": 21439
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9823086261749268,
+      "learning_rate": 4.4335183716627016e-05,
+      "loss": 1.3128,
+      "step": 21440
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.083298444747925,
+      "learning_rate": 4.433491937615649e-05,
+      "loss": 1.2836,
+      "step": 21441
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1281888484954834,
+      "learning_rate": 4.433465503568597e-05,
+      "loss": 1.177,
+      "step": 21442
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.8883135318756104,
+      "learning_rate": 4.433439069521544e-05,
+      "loss": 1.3655,
+      "step": 21443
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.4590463638305664,
+      "learning_rate": 4.433412635474492e-05,
+      "loss": 1.263,
+      "step": 21444
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.147484064102173,
+      "learning_rate": 4.4333862014274384e-05,
+      "loss": 1.3391,
+      "step": 21445
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9501267671585083,
+      "learning_rate": 4.4333597673803864e-05,
+      "loss": 1.8428,
+      "step": 21446
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2129249572753906,
+      "learning_rate": 4.433333333333334e-05,
+      "loss": 1.3935,
+      "step": 21447
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0336265563964844,
+      "learning_rate": 4.433306899286281e-05,
+      "loss": 1.3443,
+      "step": 21448
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8038976192474365,
+      "learning_rate": 4.4332804652392285e-05,
+      "loss": 1.254,
+      "step": 21449
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9691925048828125,
+      "learning_rate": 4.433254031192175e-05,
+      "loss": 1.2304,
+      "step": 21450
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.915663957595825,
+      "learning_rate": 4.433227597145123e-05,
+      "loss": 1.3453,
+      "step": 21451
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9622249603271484,
+      "learning_rate": 4.4332011630980706e-05,
+      "loss": 1.2192,
+      "step": 21452
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8282864093780518,
+      "learning_rate": 4.433174729051018e-05,
+      "loss": 1.2346,
+      "step": 21453
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.2336835861206055,
+      "learning_rate": 4.433148295003965e-05,
+      "loss": 1.3262,
+      "step": 21454
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.0654661655426025,
+      "learning_rate": 4.4331218609569127e-05,
+      "loss": 1.328,
+      "step": 21455
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8621816635131836,
+      "learning_rate": 4.43309542690986e-05,
+      "loss": 1.267,
+      "step": 21456
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.016702175140381,
+      "learning_rate": 4.433068992862808e-05,
+      "loss": 1.3658,
+      "step": 21457
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.403580665588379,
+      "learning_rate": 4.433042558815755e-05,
+      "loss": 1.2852,
+      "step": 21458
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1559972763061523,
+      "learning_rate": 4.433016124768703e-05,
+      "loss": 1.5098,
+      "step": 21459
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1260926723480225,
+      "learning_rate": 4.4329896907216494e-05,
+      "loss": 1.3315,
+      "step": 21460
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.86868155002594,
+      "learning_rate": 4.432963256674597e-05,
+      "loss": 1.2444,
+      "step": 21461
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.4889070987701416,
+      "learning_rate": 4.432936822627545e-05,
+      "loss": 1.3936,
+      "step": 21462
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6333351135253906,
+      "learning_rate": 4.4329103885804915e-05,
+      "loss": 1.2968,
+      "step": 21463
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.3062541484832764,
+      "learning_rate": 4.4328839545334396e-05,
+      "loss": 1.3475,
+      "step": 21464
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.5038046836853027,
+      "learning_rate": 4.432857520486386e-05,
+      "loss": 1.4317,
+      "step": 21465
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.804679274559021,
+      "learning_rate": 4.432831086439334e-05,
+      "loss": 1.4042,
+      "step": 21466
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.07300066947937,
+      "learning_rate": 4.4328046523922816e-05,
+      "loss": 1.3057,
+      "step": 21467
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.384260416030884,
+      "learning_rate": 4.432778218345229e-05,
+      "loss": 1.3777,
+      "step": 21468
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9357167482376099,
+      "learning_rate": 4.432751784298176e-05,
+      "loss": 1.3645,
+      "step": 21469
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.157355308532715,
+      "learning_rate": 4.432725350251124e-05,
+      "loss": 1.3629,
+      "step": 21470
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9666121006011963,
+      "learning_rate": 4.432698916204071e-05,
+      "loss": 1.2528,
+      "step": 21471
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8263179063796997,
+      "learning_rate": 4.4326724821570184e-05,
+      "loss": 1.184,
+      "step": 21472
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0965611934661865,
+      "learning_rate": 4.432646048109966e-05,
+      "loss": 1.4536,
+      "step": 21473
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.4912686347961426,
+      "learning_rate": 4.432619614062913e-05,
+      "loss": 2.1759,
+      "step": 21474
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.146097421646118,
+      "learning_rate": 4.4325931800158605e-05,
+      "loss": 1.4227,
+      "step": 21475
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6961742639541626,
+      "learning_rate": 4.432566745968808e-05,
+      "loss": 1.2062,
+      "step": 21476
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.121535539627075,
+      "learning_rate": 4.432540311921756e-05,
+      "loss": 1.3468,
+      "step": 21477
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.651404619216919,
+      "learning_rate": 4.4325138778747026e-05,
+      "loss": 1.3777,
+      "step": 21478
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7571390867233276,
+      "learning_rate": 4.4324874438276506e-05,
+      "loss": 1.1953,
+      "step": 21479
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6456202268600464,
+      "learning_rate": 4.432461009780597e-05,
+      "loss": 1.2615,
+      "step": 21480
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1645402908325195,
+      "learning_rate": 4.432434575733545e-05,
+      "loss": 1.289,
+      "step": 21481
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.072092294692993,
+      "learning_rate": 4.4324081416864927e-05,
+      "loss": 1.3244,
+      "step": 21482
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9402908086776733,
+      "learning_rate": 4.43238170763944e-05,
+      "loss": 1.9901,
+      "step": 21483
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1305277347564697,
+      "learning_rate": 4.4323552735923874e-05,
+      "loss": 1.3757,
+      "step": 21484
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3122358322143555,
+      "learning_rate": 4.432328839545335e-05,
+      "loss": 1.3293,
+      "step": 21485
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.119338274002075,
+      "learning_rate": 4.432302405498282e-05,
+      "loss": 1.3714,
+      "step": 21486
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.4368808269500732,
+      "learning_rate": 4.4322759714512294e-05,
+      "loss": 1.716,
+      "step": 21487
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2718961238861084,
+      "learning_rate": 4.432249537404177e-05,
+      "loss": 1.3291,
+      "step": 21488
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9530203342437744,
+      "learning_rate": 4.432223103357124e-05,
+      "loss": 1.3126,
+      "step": 21489
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.512791395187378,
+      "learning_rate": 4.4321966693100715e-05,
+      "loss": 1.3322,
+      "step": 21490
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2477266788482666,
+      "learning_rate": 4.432170235263019e-05,
+      "loss": 1.3655,
+      "step": 21491
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.126345634460449,
+      "learning_rate": 4.432143801215967e-05,
+      "loss": 1.4098,
+      "step": 21492
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.827991247177124,
+      "learning_rate": 4.4321173671689136e-05,
+      "loss": 1.2199,
+      "step": 21493
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7397218942642212,
+      "learning_rate": 4.4320909331218616e-05,
+      "loss": 3.4377,
+      "step": 21494
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.699957847595215,
+      "learning_rate": 4.432064499074808e-05,
+      "loss": 1.4009,
+      "step": 21495
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.33756422996521,
+      "learning_rate": 4.4320380650277557e-05,
+      "loss": 1.4266,
+      "step": 21496
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9621210098266602,
+      "learning_rate": 4.432011630980704e-05,
+      "loss": 1.2661,
+      "step": 21497
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.981835126876831,
+      "learning_rate": 4.4319851969336504e-05,
+      "loss": 1.3725,
+      "step": 21498
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8811726570129395,
+      "learning_rate": 4.4319587628865984e-05,
+      "loss": 1.2084,
+      "step": 21499
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1587841510772705,
+      "learning_rate": 4.431932328839545e-05,
+      "loss": 1.3445,
+      "step": 21500
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.5161705017089844,
+      "learning_rate": 4.431905894792493e-05,
+      "loss": 1.5176,
+      "step": 21501
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.883268117904663,
+      "learning_rate": 4.4318794607454405e-05,
+      "loss": 1.2852,
+      "step": 21502
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9301804304122925,
+      "learning_rate": 4.431853026698388e-05,
+      "loss": 1.3599,
+      "step": 21503
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.5120869874954224,
+      "learning_rate": 4.431826592651335e-05,
+      "loss": 1.1005,
+      "step": 21504
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.373087167739868,
+      "learning_rate": 4.4318001586042825e-05,
+      "loss": 1.3343,
+      "step": 21505
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9543513059616089,
+      "learning_rate": 4.43177372455723e-05,
+      "loss": 1.2748,
+      "step": 21506
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6584978103637695,
+      "learning_rate": 4.431747290510177e-05,
+      "loss": 1.1598,
+      "step": 21507
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9848413467407227,
+      "learning_rate": 4.4317208564631246e-05,
+      "loss": 1.3833,
+      "step": 21508
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0865814685821533,
+      "learning_rate": 4.431694422416072e-05,
+      "loss": 1.3498,
+      "step": 21509
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3712170124053955,
+      "learning_rate": 4.431667988369019e-05,
+      "loss": 1.7264,
+      "step": 21510
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0927155017852783,
+      "learning_rate": 4.431641554321967e-05,
+      "loss": 1.1007,
+      "step": 21511
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0261008739471436,
+      "learning_rate": 4.431615120274915e-05,
+      "loss": 1.4137,
+      "step": 21512
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8163727521896362,
+      "learning_rate": 4.4315886862278614e-05,
+      "loss": 1.2392,
+      "step": 21513
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.437192678451538,
+      "learning_rate": 4.4315622521808094e-05,
+      "loss": 1.5228,
+      "step": 21514
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3751163482666016,
+      "learning_rate": 4.431535818133756e-05,
+      "loss": 1.2316,
+      "step": 21515
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0260586738586426,
+      "learning_rate": 4.431509384086704e-05,
+      "loss": 1.246,
+      "step": 21516
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2705724239349365,
+      "learning_rate": 4.4314829500396515e-05,
+      "loss": 1.5863,
+      "step": 21517
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8385727405548096,
+      "learning_rate": 4.431456515992599e-05,
+      "loss": 1.3222,
+      "step": 21518
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.2641382217407227,
+      "learning_rate": 4.431430081945546e-05,
+      "loss": 1.5079,
+      "step": 21519
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.031344413757324,
+      "learning_rate": 4.4314036478984936e-05,
+      "loss": 1.5131,
+      "step": 21520
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.108600616455078,
+      "learning_rate": 4.431377213851441e-05,
+      "loss": 1.3022,
+      "step": 21521
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6266218423843384,
+      "learning_rate": 4.431350779804388e-05,
+      "loss": 1.1474,
+      "step": 21522
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2102413177490234,
+      "learning_rate": 4.4313243457573357e-05,
+      "loss": 1.2621,
+      "step": 21523
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0077784061431885,
+      "learning_rate": 4.431297911710283e-05,
+      "loss": 1.2924,
+      "step": 21524
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3628945350646973,
+      "learning_rate": 4.4312714776632304e-05,
+      "loss": 1.2653,
+      "step": 21525
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8274208307266235,
+      "learning_rate": 4.431245043616178e-05,
+      "loss": 1.2336,
+      "step": 21526
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0270512104034424,
+      "learning_rate": 4.431218609569126e-05,
+      "loss": 1.2154,
+      "step": 21527
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8110674619674683,
+      "learning_rate": 4.4311921755220724e-05,
+      "loss": 2.2511,
+      "step": 21528
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.125020980834961,
+      "learning_rate": 4.4311657414750205e-05,
+      "loss": 1.3781,
+      "step": 21529
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.514556407928467,
+      "learning_rate": 4.431139307427967e-05,
+      "loss": 1.4018,
+      "step": 21530
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.5011260509490967,
+      "learning_rate": 4.431112873380915e-05,
+      "loss": 1.4279,
+      "step": 21531
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7510595321655273,
+      "learning_rate": 4.4310864393338625e-05,
+      "loss": 1.2658,
+      "step": 21532
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9155097007751465,
+      "learning_rate": 4.431060005286809e-05,
+      "loss": 1.1737,
+      "step": 21533
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3558571338653564,
+      "learning_rate": 4.431033571239757e-05,
+      "loss": 1.3277,
+      "step": 21534
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.805281400680542,
+      "learning_rate": 4.431007137192704e-05,
+      "loss": 1.2283,
+      "step": 21535
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.203598737716675,
+      "learning_rate": 4.430980703145652e-05,
+      "loss": 1.3278,
+      "step": 21536
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.784334897994995,
+      "learning_rate": 4.430954269098599e-05,
+      "loss": 1.5368,
+      "step": 21537
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7633123397827148,
+      "learning_rate": 4.430927835051547e-05,
+      "loss": 1.1896,
+      "step": 21538
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.452225923538208,
+      "learning_rate": 4.430901401004494e-05,
+      "loss": 2.1252,
+      "step": 21539
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.33489727973938,
+      "learning_rate": 4.4308749669574414e-05,
+      "loss": 1.5202,
+      "step": 21540
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7075986862182617,
+      "learning_rate": 4.430848532910389e-05,
+      "loss": 1.8862,
+      "step": 21541
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.893247365951538,
+      "learning_rate": 4.430822098863336e-05,
+      "loss": 1.8733,
+      "step": 21542
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.507533311843872,
+      "learning_rate": 4.4307956648162835e-05,
+      "loss": 1.5043,
+      "step": 21543
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.518923282623291,
+      "learning_rate": 4.430769230769231e-05,
+      "loss": 1.4666,
+      "step": 21544
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9977682828903198,
+      "learning_rate": 4.430742796722178e-05,
+      "loss": 1.2313,
+      "step": 21545
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7124289274215698,
+      "learning_rate": 4.4307163626751255e-05,
+      "loss": 2.8699,
+      "step": 21546
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0750513076782227,
+      "learning_rate": 4.4306899286280736e-05,
+      "loss": 1.3971,
+      "step": 21547
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.939733862876892,
+      "learning_rate": 4.43066349458102e-05,
+      "loss": 1.259,
+      "step": 21548
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.5791025161743164,
+      "learning_rate": 4.430637060533968e-05,
+      "loss": 1.3249,
+      "step": 21549
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.494297504425049,
+      "learning_rate": 4.430610626486915e-05,
+      "loss": 1.4337,
+      "step": 21550
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7235509157180786,
+      "learning_rate": 4.430584192439863e-05,
+      "loss": 1.1472,
+      "step": 21551
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.6011881828308105,
+      "learning_rate": 4.4305577583928104e-05,
+      "loss": 1.2779,
+      "step": 21552
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.663344144821167,
+      "learning_rate": 4.430531324345758e-05,
+      "loss": 1.2167,
+      "step": 21553
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.017483949661255,
+      "learning_rate": 4.430504890298705e-05,
+      "loss": 1.3795,
+      "step": 21554
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.017514944076538,
+      "learning_rate": 4.4304784562516524e-05,
+      "loss": 1.2635,
+      "step": 21555
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3778772354125977,
+      "learning_rate": 4.4304520222046e-05,
+      "loss": 1.409,
+      "step": 21556
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9196338653564453,
+      "learning_rate": 4.430425588157547e-05,
+      "loss": 2.0956,
+      "step": 21557
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.124774694442749,
+      "learning_rate": 4.4303991541104945e-05,
+      "loss": 1.2955,
+      "step": 21558
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9381624460220337,
+      "learning_rate": 4.430372720063442e-05,
+      "loss": 1.3165,
+      "step": 21559
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.5127034187316895,
+      "learning_rate": 4.430346286016389e-05,
+      "loss": 1.2952,
+      "step": 21560
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.81917405128479,
+      "learning_rate": 4.4303198519693366e-05,
+      "loss": 1.202,
+      "step": 21561
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.956056833267212,
+      "learning_rate": 4.4302934179222846e-05,
+      "loss": 1.1846,
+      "step": 21562
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.194915533065796,
+      "learning_rate": 4.430266983875231e-05,
+      "loss": 1.0957,
+      "step": 21563
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9137951135635376,
+      "learning_rate": 4.430240549828179e-05,
+      "loss": 1.3346,
+      "step": 21564
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1114187240600586,
+      "learning_rate": 4.430214115781126e-05,
+      "loss": 1.3823,
+      "step": 21565
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.433838367462158,
+      "learning_rate": 4.430187681734074e-05,
+      "loss": 1.2134,
+      "step": 21566
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1385838985443115,
+      "learning_rate": 4.4301612476870214e-05,
+      "loss": 1.2694,
+      "step": 21567
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9544850587844849,
+      "learning_rate": 4.430134813639968e-05,
+      "loss": 1.3153,
+      "step": 21568
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9066392183303833,
+      "learning_rate": 4.430108379592916e-05,
+      "loss": 1.2166,
+      "step": 21569
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1035068035125732,
+      "learning_rate": 4.430081945545863e-05,
+      "loss": 1.1587,
+      "step": 21570
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.14430832862854,
+      "learning_rate": 4.430055511498811e-05,
+      "loss": 1.1137,
+      "step": 21571
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.5818722248077393,
+      "learning_rate": 4.430029077451758e-05,
+      "loss": 1.1483,
+      "step": 21572
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1114256381988525,
+      "learning_rate": 4.4300026434047055e-05,
+      "loss": 1.2343,
+      "step": 21573
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9061439037322998,
+      "learning_rate": 4.429976209357653e-05,
+      "loss": 1.2453,
+      "step": 21574
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0395214557647705,
+      "learning_rate": 4.4299497753106e-05,
+      "loss": 1.3686,
+      "step": 21575
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0283710956573486,
+      "learning_rate": 4.4299233412635476e-05,
+      "loss": 1.3519,
+      "step": 21576
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9918571710586548,
+      "learning_rate": 4.429896907216495e-05,
+      "loss": 1.3064,
+      "step": 21577
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8949438333511353,
+      "learning_rate": 4.429870473169442e-05,
+      "loss": 1.2476,
+      "step": 21578
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9715520143508911,
+      "learning_rate": 4.42984403912239e-05,
+      "loss": 1.2735,
+      "step": 21579
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.820490837097168,
+      "learning_rate": 4.429817605075337e-05,
+      "loss": 1.1823,
+      "step": 21580
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9571492671966553,
+      "learning_rate": 4.4297911710282844e-05,
+      "loss": 1.3129,
+      "step": 21581
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.129627227783203,
+      "learning_rate": 4.4297647369812324e-05,
+      "loss": 1.2248,
+      "step": 21582
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.555004835128784,
+      "learning_rate": 4.429738302934179e-05,
+      "loss": 1.3664,
+      "step": 21583
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.7638301849365234,
+      "learning_rate": 4.429711868887127e-05,
+      "loss": 1.6221,
+      "step": 21584
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.047692060470581,
+      "learning_rate": 4.429685434840074e-05,
+      "loss": 1.3261,
+      "step": 21585
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8126857280731201,
+      "learning_rate": 4.429659000793022e-05,
+      "loss": 1.2597,
+      "step": 21586
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1544601917266846,
+      "learning_rate": 4.429632566745969e-05,
+      "loss": 1.2671,
+      "step": 21587
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7526293992996216,
+      "learning_rate": 4.4296061326989166e-05,
+      "loss": 1.1756,
+      "step": 21588
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.5416641235351562,
+      "learning_rate": 4.429579698651864e-05,
+      "loss": 1.3536,
+      "step": 21589
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.32600998878479,
+      "learning_rate": 4.429553264604811e-05,
+      "loss": 1.3635,
+      "step": 21590
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.4876930713653564,
+      "learning_rate": 4.4295268305577586e-05,
+      "loss": 1.6209,
+      "step": 21591
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7000755071640015,
+      "learning_rate": 4.429500396510706e-05,
+      "loss": 1.3409,
+      "step": 21592
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.875990390777588,
+      "learning_rate": 4.4294739624636534e-05,
+      "loss": 1.2312,
+      "step": 21593
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.4765889644622803,
+      "learning_rate": 4.429447528416601e-05,
+      "loss": 1.2991,
+      "step": 21594
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.5066282749176025,
+      "learning_rate": 4.429421094369548e-05,
+      "loss": 1.3781,
+      "step": 21595
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.8265445232391357,
+      "learning_rate": 4.4293946603224954e-05,
+      "loss": 1.3249,
+      "step": 21596
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6536221504211426,
+      "learning_rate": 4.4293682262754435e-05,
+      "loss": 1.2393,
+      "step": 21597
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.4369921684265137,
+      "learning_rate": 4.42934179222839e-05,
+      "loss": 1.4077,
+      "step": 21598
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2616145610809326,
+      "learning_rate": 4.429315358181338e-05,
+      "loss": 1.422,
+      "step": 21599
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.874455451965332,
+      "learning_rate": 4.429288924134285e-05,
+      "loss": 1.3549,
+      "step": 21600
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7821944952011108,
+      "learning_rate": 4.429262490087233e-05,
+      "loss": 1.1311,
+      "step": 21601
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.61393404006958,
+      "learning_rate": 4.42923605604018e-05,
+      "loss": 1.0966,
+      "step": 21602
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1593966484069824,
+      "learning_rate": 4.4292096219931276e-05,
+      "loss": 1.3639,
+      "step": 21603
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.745931625366211,
+      "learning_rate": 4.429183187946075e-05,
+      "loss": 1.1653,
+      "step": 21604
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.747539520263672,
+      "learning_rate": 4.4291567538990216e-05,
+      "loss": 1.1484,
+      "step": 21605
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2972943782806396,
+      "learning_rate": 4.42913031985197e-05,
+      "loss": 1.4581,
+      "step": 21606
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1556105613708496,
+      "learning_rate": 4.429103885804917e-05,
+      "loss": 1.3825,
+      "step": 21607
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7808327674865723,
+      "learning_rate": 4.4290774517578644e-05,
+      "loss": 1.2113,
+      "step": 21608
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8816876411437988,
+      "learning_rate": 4.429051017710812e-05,
+      "loss": 1.2639,
+      "step": 21609
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9628440141677856,
+      "learning_rate": 4.429024583663759e-05,
+      "loss": 1.3311,
+      "step": 21610
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8417387008666992,
+      "learning_rate": 4.4289981496167065e-05,
+      "loss": 1.3219,
+      "step": 21611
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.4655370712280273,
+      "learning_rate": 4.4289717155696545e-05,
+      "loss": 1.2061,
+      "step": 21612
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7698249816894531,
+      "learning_rate": 4.428945281522601e-05,
+      "loss": 1.3323,
+      "step": 21613
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.423649549484253,
+      "learning_rate": 4.4289188474755485e-05,
+      "loss": 1.4486,
+      "step": 21614
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.646674633026123,
+      "learning_rate": 4.428892413428496e-05,
+      "loss": 1.2862,
+      "step": 21615
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7839634418487549,
+      "learning_rate": 4.428865979381443e-05,
+      "loss": 1.2847,
+      "step": 21616
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3260483741760254,
+      "learning_rate": 4.428839545334391e-05,
+      "loss": 1.3609,
+      "step": 21617
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.417729139328003,
+      "learning_rate": 4.428813111287338e-05,
+      "loss": 1.3561,
+      "step": 21618
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9931834936141968,
+      "learning_rate": 4.428786677240286e-05,
+      "loss": 1.2553,
+      "step": 21619
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3837945461273193,
+      "learning_rate": 4.428760243193233e-05,
+      "loss": 1.3614,
+      "step": 21620
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3769195079803467,
+      "learning_rate": 4.428733809146181e-05,
+      "loss": 1.4097,
+      "step": 21621
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7244861125946045,
+      "learning_rate": 4.428707375099128e-05,
+      "loss": 1.2425,
+      "step": 21622
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.410205841064453,
+      "learning_rate": 4.4286809410520754e-05,
+      "loss": 1.4365,
+      "step": 21623
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.243624448776245,
+      "learning_rate": 4.428654507005023e-05,
+      "loss": 1.3316,
+      "step": 21624
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.735794186592102,
+      "learning_rate": 4.42862807295797e-05,
+      "loss": 1.2018,
+      "step": 21625
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7557413578033447,
+      "learning_rate": 4.4286016389109175e-05,
+      "loss": 1.2669,
+      "step": 21626
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.7266101837158203,
+      "learning_rate": 4.428575204863865e-05,
+      "loss": 1.2724,
+      "step": 21627
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.054187536239624,
+      "learning_rate": 4.428548770816812e-05,
+      "loss": 1.3012,
+      "step": 21628
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.036259889602661,
+      "learning_rate": 4.4285223367697596e-05,
+      "loss": 1.3331,
+      "step": 21629
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1414315700531006,
+      "learning_rate": 4.428495902722707e-05,
+      "loss": 1.5772,
+      "step": 21630
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9592069387435913,
+      "learning_rate": 4.428469468675654e-05,
+      "loss": 1.3843,
+      "step": 21631
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.434080123901367,
+      "learning_rate": 4.428443034628602e-05,
+      "loss": 1.3747,
+      "step": 21632
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.337554693222046,
+      "learning_rate": 4.428416600581549e-05,
+      "loss": 1.3922,
+      "step": 21633
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.728553056716919,
+      "learning_rate": 4.428390166534497e-05,
+      "loss": 1.3069,
+      "step": 21634
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.203383684158325,
+      "learning_rate": 4.428363732487444e-05,
+      "loss": 1.3567,
+      "step": 21635
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3465514183044434,
+      "learning_rate": 4.428337298440392e-05,
+      "loss": 1.4331,
+      "step": 21636
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7678438425064087,
+      "learning_rate": 4.428310864393339e-05,
+      "loss": 1.2702,
+      "step": 21637
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.184507369995117,
+      "learning_rate": 4.4282844303462865e-05,
+      "loss": 1.3331,
+      "step": 21638
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1284544467926025,
+      "learning_rate": 4.428257996299234e-05,
+      "loss": 1.3678,
+      "step": 21639
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6605890989303589,
+      "learning_rate": 4.4282315622521805e-05,
+      "loss": 1.2032,
+      "step": 21640
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8580487966537476,
+      "learning_rate": 4.4282051282051285e-05,
+      "loss": 1.2688,
+      "step": 21641
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1194705963134766,
+      "learning_rate": 4.428178694158076e-05,
+      "loss": 1.4718,
+      "step": 21642
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6120450496673584,
+      "learning_rate": 4.428152260111023e-05,
+      "loss": 1.1295,
+      "step": 21643
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8799036741256714,
+      "learning_rate": 4.4281258260639706e-05,
+      "loss": 1.2894,
+      "step": 21644
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8435289859771729,
+      "learning_rate": 4.428099392016918e-05,
+      "loss": 1.21,
+      "step": 21645
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.170196056365967,
+      "learning_rate": 4.428072957969865e-05,
+      "loss": 1.365,
+      "step": 21646
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2391204833984375,
+      "learning_rate": 4.4280465239228134e-05,
+      "loss": 1.3857,
+      "step": 21647
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8899998664855957,
+      "learning_rate": 4.42802008987576e-05,
+      "loss": 1.2946,
+      "step": 21648
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3915014266967773,
+      "learning_rate": 4.427993655828708e-05,
+      "loss": 1.2619,
+      "step": 21649
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7712056636810303,
+      "learning_rate": 4.427967221781655e-05,
+      "loss": 1.1656,
+      "step": 21650
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.685495138168335,
+      "learning_rate": 4.427940787734602e-05,
+      "loss": 1.1929,
+      "step": 21651
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.491816520690918,
+      "learning_rate": 4.42791435368755e-05,
+      "loss": 1.4322,
+      "step": 21652
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.574251174926758,
+      "learning_rate": 4.427887919640497e-05,
+      "loss": 1.6038,
+      "step": 21653
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8444862365722656,
+      "learning_rate": 4.427861485593445e-05,
+      "loss": 1.3281,
+      "step": 21654
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8975987434387207,
+      "learning_rate": 4.4278350515463915e-05,
+      "loss": 1.3062,
+      "step": 21655
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7494572401046753,
+      "learning_rate": 4.4278086174993396e-05,
+      "loss": 1.4065,
+      "step": 21656
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8921464681625366,
+      "learning_rate": 4.427782183452287e-05,
+      "loss": 1.3396,
+      "step": 21657
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.371994733810425,
+      "learning_rate": 4.427755749405234e-05,
+      "loss": 1.3886,
+      "step": 21658
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.5895252227783203,
+      "learning_rate": 4.4277293153581816e-05,
+      "loss": 1.5424,
+      "step": 21659
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9761627912521362,
+      "learning_rate": 4.427702881311129e-05,
+      "loss": 1.0989,
+      "step": 21660
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9433420896530151,
+      "learning_rate": 4.4276764472640764e-05,
+      "loss": 1.1919,
+      "step": 21661
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.5727511644363403,
+      "learning_rate": 4.427650013217024e-05,
+      "loss": 1.8877,
+      "step": 21662
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.948717713356018,
+      "learning_rate": 4.427623579169971e-05,
+      "loss": 1.1464,
+      "step": 21663
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.674344062805176,
+      "learning_rate": 4.4275971451229184e-05,
+      "loss": 1.2353,
+      "step": 21664
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7624900341033936,
+      "learning_rate": 4.427570711075866e-05,
+      "loss": 1.2849,
+      "step": 21665
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8850563764572144,
+      "learning_rate": 4.427544277028813e-05,
+      "loss": 1.3181,
+      "step": 21666
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7505104541778564,
+      "learning_rate": 4.427517842981761e-05,
+      "loss": 1.3353,
+      "step": 21667
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9240550994873047,
+      "learning_rate": 4.427491408934708e-05,
+      "loss": 1.2307,
+      "step": 21668
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9862815141677856,
+      "learning_rate": 4.427464974887656e-05,
+      "loss": 1.3817,
+      "step": 21669
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.9602229595184326,
+      "learning_rate": 4.4274385408406026e-05,
+      "loss": 1.5539,
+      "step": 21670
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.466888427734375,
+      "learning_rate": 4.4274121067935506e-05,
+      "loss": 1.5104,
+      "step": 21671
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8142181634902954,
+      "learning_rate": 4.427385672746498e-05,
+      "loss": 1.6991,
+      "step": 21672
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.300877332687378,
+      "learning_rate": 4.427359238699445e-05,
+      "loss": 1.4054,
+      "step": 21673
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.891296625137329,
+      "learning_rate": 4.427332804652393e-05,
+      "loss": 1.3981,
+      "step": 21674
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.9401047229766846,
+      "learning_rate": 4.42730637060534e-05,
+      "loss": 1.3878,
+      "step": 21675
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2123968601226807,
+      "learning_rate": 4.4272799365582874e-05,
+      "loss": 1.4368,
+      "step": 21676
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.4379233121871948,
+      "learning_rate": 4.427253502511235e-05,
+      "loss": 1.0821,
+      "step": 21677
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7051314115524292,
+      "learning_rate": 4.427227068464182e-05,
+      "loss": 1.2367,
+      "step": 21678
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6901942491531372,
+      "learning_rate": 4.4272006344171295e-05,
+      "loss": 1.2385,
+      "step": 21679
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9601038694381714,
+      "learning_rate": 4.427174200370077e-05,
+      "loss": 1.2978,
+      "step": 21680
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9047377109527588,
+      "learning_rate": 4.427147766323024e-05,
+      "loss": 1.3332,
+      "step": 21681
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.0333197116851807,
+      "learning_rate": 4.427121332275972e-05,
+      "loss": 1.4326,
+      "step": 21682
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.081911325454712,
+      "learning_rate": 4.427094898228919e-05,
+      "loss": 1.238,
+      "step": 21683
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.681767225265503,
+      "learning_rate": 4.427068464181867e-05,
+      "loss": 1.1956,
+      "step": 21684
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.832504391670227,
+      "learning_rate": 4.4270420301348136e-05,
+      "loss": 1.2616,
+      "step": 21685
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9034496545791626,
+      "learning_rate": 4.427015596087761e-05,
+      "loss": 1.3478,
+      "step": 21686
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.355602741241455,
+      "learning_rate": 4.426989162040709e-05,
+      "loss": 1.3992,
+      "step": 21687
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9562801122665405,
+      "learning_rate": 4.426962727993656e-05,
+      "loss": 1.2866,
+      "step": 21688
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.691290855407715,
+      "learning_rate": 4.426936293946604e-05,
+      "loss": 1.4226,
+      "step": 21689
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1957545280456543,
+      "learning_rate": 4.4269098598995504e-05,
+      "loss": 1.2953,
+      "step": 21690
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.6036624908447266,
+      "learning_rate": 4.4268834258524984e-05,
+      "loss": 1.2428,
+      "step": 21691
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.608978509902954,
+      "learning_rate": 4.426856991805446e-05,
+      "loss": 1.5804,
+      "step": 21692
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.344294548034668,
+      "learning_rate": 4.426830557758393e-05,
+      "loss": 1.4632,
+      "step": 21693
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0349886417388916,
+      "learning_rate": 4.4268041237113405e-05,
+      "loss": 1.3578,
+      "step": 21694
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2921111583709717,
+      "learning_rate": 4.426777689664288e-05,
+      "loss": 1.3961,
+      "step": 21695
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8172729015350342,
+      "learning_rate": 4.426751255617235e-05,
+      "loss": 1.1056,
+      "step": 21696
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9643714427947998,
+      "learning_rate": 4.4267248215701826e-05,
+      "loss": 1.3182,
+      "step": 21697
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0835375785827637,
+      "learning_rate": 4.42669838752313e-05,
+      "loss": 1.4921,
+      "step": 21698
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8677462339401245,
+      "learning_rate": 4.426671953476077e-05,
+      "loss": 1.1433,
+      "step": 21699
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6806071996688843,
+      "learning_rate": 4.4266455194290246e-05,
+      "loss": 1.328,
+      "step": 21700
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.862876057624817,
+      "learning_rate": 4.426619085381972e-05,
+      "loss": 1.3035,
+      "step": 21701
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.045656442642212,
+      "learning_rate": 4.42659265133492e-05,
+      "loss": 1.2683,
+      "step": 21702
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0472590923309326,
+      "learning_rate": 4.426566217287867e-05,
+      "loss": 1.2514,
+      "step": 21703
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2126686573028564,
+      "learning_rate": 4.426539783240815e-05,
+      "loss": 1.3413,
+      "step": 21704
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9175572395324707,
+      "learning_rate": 4.4265133491937614e-05,
+      "loss": 1.2724,
+      "step": 21705
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.051462411880493,
+      "learning_rate": 4.4264869151467095e-05,
+      "loss": 1.1395,
+      "step": 21706
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.940396785736084,
+      "learning_rate": 4.426460481099657e-05,
+      "loss": 1.4497,
+      "step": 21707
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.369723081588745,
+      "learning_rate": 4.426434047052604e-05,
+      "loss": 1.2411,
+      "step": 21708
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7106964588165283,
+      "learning_rate": 4.4264076130055515e-05,
+      "loss": 1.7864,
+      "step": 21709
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.453035593032837,
+      "learning_rate": 4.426381178958499e-05,
+      "loss": 1.3598,
+      "step": 21710
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.474172353744507,
+      "learning_rate": 4.426354744911446e-05,
+      "loss": 1.3708,
+      "step": 21711
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.261889696121216,
+      "learning_rate": 4.4263283108643936e-05,
+      "loss": 1.3579,
+      "step": 21712
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.229452133178711,
+      "learning_rate": 4.426301876817341e-05,
+      "loss": 1.2501,
+      "step": 21713
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0956437587738037,
+      "learning_rate": 4.426275442770288e-05,
+      "loss": 1.3631,
+      "step": 21714
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7598755359649658,
+      "learning_rate": 4.426249008723236e-05,
+      "loss": 1.1647,
+      "step": 21715
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.953642725944519,
+      "learning_rate": 4.426222574676183e-05,
+      "loss": 1.1746,
+      "step": 21716
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.9192678928375244,
+      "learning_rate": 4.426196140629131e-05,
+      "loss": 1.446,
+      "step": 21717
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.407076835632324,
+      "learning_rate": 4.426169706582078e-05,
+      "loss": 1.4495,
+      "step": 21718
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9333463907241821,
+      "learning_rate": 4.426143272535026e-05,
+      "loss": 1.3047,
+      "step": 21719
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.9595298767089844,
+      "learning_rate": 4.4261168384879725e-05,
+      "loss": 1.4002,
+      "step": 21720
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6872401237487793,
+      "learning_rate": 4.4260904044409205e-05,
+      "loss": 1.9276,
+      "step": 21721
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.3560919761657715,
+      "learning_rate": 4.426063970393868e-05,
+      "loss": 1.379,
+      "step": 21722
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8083858489990234,
+      "learning_rate": 4.4260375363468145e-05,
+      "loss": 1.2801,
+      "step": 21723
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8685662746429443,
+      "learning_rate": 4.4260111022997626e-05,
+      "loss": 1.297,
+      "step": 21724
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8492573499679565,
+      "learning_rate": 4.425984668252709e-05,
+      "loss": 1.3357,
+      "step": 21725
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.3456170558929443,
+      "learning_rate": 4.425958234205657e-05,
+      "loss": 1.3352,
+      "step": 21726
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0734262466430664,
+      "learning_rate": 4.4259318001586046e-05,
+      "loss": 1.2515,
+      "step": 21727
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.037491798400879,
+      "learning_rate": 4.425905366111552e-05,
+      "loss": 1.4729,
+      "step": 21728
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7594982385635376,
+      "learning_rate": 4.4258789320644993e-05,
+      "loss": 1.4012,
+      "step": 21729
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9161943197250366,
+      "learning_rate": 4.425852498017447e-05,
+      "loss": 1.2265,
+      "step": 21730
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7065534591674805,
+      "learning_rate": 4.425826063970394e-05,
+      "loss": 1.2218,
+      "step": 21731
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.031975746154785,
+      "learning_rate": 4.4257996299233414e-05,
+      "loss": 1.2848,
+      "step": 21732
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9890791177749634,
+      "learning_rate": 4.425773195876289e-05,
+      "loss": 1.3685,
+      "step": 21733
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9816055297851562,
+      "learning_rate": 4.425746761829236e-05,
+      "loss": 1.2996,
+      "step": 21734
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1576309204101562,
+      "learning_rate": 4.4257203277821835e-05,
+      "loss": 1.4229,
+      "step": 21735
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.476629376411438,
+      "learning_rate": 4.425693893735131e-05,
+      "loss": 1.2392,
+      "step": 21736
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9311671257019043,
+      "learning_rate": 4.425667459688079e-05,
+      "loss": 1.2604,
+      "step": 21737
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6295099258422852,
+      "learning_rate": 4.4256410256410256e-05,
+      "loss": 1.2295,
+      "step": 21738
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2121291160583496,
+      "learning_rate": 4.4256145915939736e-05,
+      "loss": 1.2442,
+      "step": 21739
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.6410562992095947,
+      "learning_rate": 4.42558815754692e-05,
+      "loss": 1.2566,
+      "step": 21740
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.978446125984192,
+      "learning_rate": 4.425561723499868e-05,
+      "loss": 1.2282,
+      "step": 21741
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8033727407455444,
+      "learning_rate": 4.425535289452816e-05,
+      "loss": 1.9663,
+      "step": 21742
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.3165202140808105,
+      "learning_rate": 4.425508855405763e-05,
+      "loss": 1.2846,
+      "step": 21743
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8420017957687378,
+      "learning_rate": 4.4254824213587104e-05,
+      "loss": 1.1996,
+      "step": 21744
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5356898307800293,
+      "learning_rate": 4.425455987311658e-05,
+      "loss": 1.4361,
+      "step": 21745
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2697763442993164,
+      "learning_rate": 4.425429553264605e-05,
+      "loss": 1.2514,
+      "step": 21746
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.6146998405456543,
+      "learning_rate": 4.4254031192175525e-05,
+      "loss": 1.4448,
+      "step": 21747
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9158408641815186,
+      "learning_rate": 4.4253766851705e-05,
+      "loss": 1.2338,
+      "step": 21748
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.9531686305999756,
+      "learning_rate": 4.425350251123447e-05,
+      "loss": 1.4035,
+      "step": 21749
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7397576570510864,
+      "learning_rate": 4.4253238170763945e-05,
+      "loss": 1.3196,
+      "step": 21750
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.0238261222839355,
+      "learning_rate": 4.425297383029342e-05,
+      "loss": 1.4004,
+      "step": 21751
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.347085952758789,
+      "learning_rate": 4.42527094898229e-05,
+      "loss": 1.2158,
+      "step": 21752
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1431467533111572,
+      "learning_rate": 4.4252445149352366e-05,
+      "loss": 1.2723,
+      "step": 21753
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.376235008239746,
+      "learning_rate": 4.4252180808881846e-05,
+      "loss": 1.3558,
+      "step": 21754
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5429656505584717,
+      "learning_rate": 4.425191646841131e-05,
+      "loss": 1.3922,
+      "step": 21755
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4876177310943604,
+      "learning_rate": 4.4251652127940793e-05,
+      "loss": 1.4128,
+      "step": 21756
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8417439460754395,
+      "learning_rate": 4.425138778747027e-05,
+      "loss": 1.1841,
+      "step": 21757
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1767964363098145,
+      "learning_rate": 4.4251123446999734e-05,
+      "loss": 1.3837,
+      "step": 21758
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.111597776412964,
+      "learning_rate": 4.4250859106529214e-05,
+      "loss": 1.3142,
+      "step": 21759
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9744255542755127,
+      "learning_rate": 4.425059476605868e-05,
+      "loss": 1.3075,
+      "step": 21760
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9429028034210205,
+      "learning_rate": 4.425033042558816e-05,
+      "loss": 2.0488,
+      "step": 21761
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.6067686080932617,
+      "learning_rate": 4.4250066085117635e-05,
+      "loss": 1.3429,
+      "step": 21762
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.593589425086975,
+      "learning_rate": 4.424980174464711e-05,
+      "loss": 1.3037,
+      "step": 21763
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4153833389282227,
+      "learning_rate": 4.424953740417658e-05,
+      "loss": 1.4838,
+      "step": 21764
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.5833489894866943,
+      "learning_rate": 4.4249273063706056e-05,
+      "loss": 1.6188,
+      "step": 21765
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6832681894302368,
+      "learning_rate": 4.424900872323553e-05,
+      "loss": 1.226,
+      "step": 21766
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1615161895751953,
+      "learning_rate": 4.424874438276501e-05,
+      "loss": 1.2951,
+      "step": 21767
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0330421924591064,
+      "learning_rate": 4.4248480042294476e-05,
+      "loss": 1.2352,
+      "step": 21768
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4027469158172607,
+      "learning_rate": 4.424821570182395e-05,
+      "loss": 1.3834,
+      "step": 21769
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.7550644874572754,
+      "learning_rate": 4.4247951361353423e-05,
+      "loss": 1.3012,
+      "step": 21770
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0146608352661133,
+      "learning_rate": 4.42476870208829e-05,
+      "loss": 1.2666,
+      "step": 21771
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4627511501312256,
+      "learning_rate": 4.424742268041238e-05,
+      "loss": 1.38,
+      "step": 21772
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.354415416717529,
+      "learning_rate": 4.4247158339941844e-05,
+      "loss": 1.3466,
+      "step": 21773
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4559762477874756,
+      "learning_rate": 4.4246893999471324e-05,
+      "loss": 1.3032,
+      "step": 21774
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.3403828144073486,
+      "learning_rate": 4.424662965900079e-05,
+      "loss": 1.1296,
+      "step": 21775
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.056950569152832,
+      "learning_rate": 4.424636531853027e-05,
+      "loss": 1.2535,
+      "step": 21776
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0461387634277344,
+      "learning_rate": 4.4246100978059745e-05,
+      "loss": 1.1448,
+      "step": 21777
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9064137935638428,
+      "learning_rate": 4.424583663758922e-05,
+      "loss": 1.1297,
+      "step": 21778
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8535435199737549,
+      "learning_rate": 4.424557229711869e-05,
+      "loss": 1.2774,
+      "step": 21779
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8090269565582275,
+      "learning_rate": 4.4245307956648166e-05,
+      "loss": 1.1238,
+      "step": 21780
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5430469512939453,
+      "learning_rate": 4.424504361617764e-05,
+      "loss": 1.2409,
+      "step": 21781
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.5510931015014648,
+      "learning_rate": 4.424477927570711e-05,
+      "loss": 1.2283,
+      "step": 21782
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7000995874404907,
+      "learning_rate": 4.4244514935236587e-05,
+      "loss": 1.1779,
+      "step": 21783
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.158702850341797,
+      "learning_rate": 4.424425059476606e-05,
+      "loss": 1.3232,
+      "step": 21784
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6485942602157593,
+      "learning_rate": 4.4243986254295534e-05,
+      "loss": 1.176,
+      "step": 21785
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.8377797603607178,
+      "learning_rate": 4.424372191382501e-05,
+      "loss": 1.2948,
+      "step": 21786
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.081472873687744,
+      "learning_rate": 4.424345757335449e-05,
+      "loss": 1.315,
+      "step": 21787
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4837303161621094,
+      "learning_rate": 4.4243193232883954e-05,
+      "loss": 1.8963,
+      "step": 21788
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0470056533813477,
+      "learning_rate": 4.4242928892413435e-05,
+      "loss": 1.2422,
+      "step": 21789
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.592667818069458,
+      "learning_rate": 4.42426645519429e-05,
+      "loss": 1.1682,
+      "step": 21790
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.903477430343628,
+      "learning_rate": 4.424240021147238e-05,
+      "loss": 1.3285,
+      "step": 21791
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.836380124092102,
+      "learning_rate": 4.4242135871001856e-05,
+      "loss": 1.3773,
+      "step": 21792
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7891943454742432,
+      "learning_rate": 4.424187153053133e-05,
+      "loss": 1.3,
+      "step": 21793
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7572206258773804,
+      "learning_rate": 4.42416071900608e-05,
+      "loss": 1.1207,
+      "step": 21794
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.386526107788086,
+      "learning_rate": 4.424134284959027e-05,
+      "loss": 1.4757,
+      "step": 21795
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.562654972076416,
+      "learning_rate": 4.424107850911975e-05,
+      "loss": 1.3839,
+      "step": 21796
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8099428415298462,
+      "learning_rate": 4.424081416864922e-05,
+      "loss": 1.1501,
+      "step": 21797
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9410381317138672,
+      "learning_rate": 4.42405498281787e-05,
+      "loss": 1.2731,
+      "step": 21798
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7503175735473633,
+      "learning_rate": 4.424028548770817e-05,
+      "loss": 1.2826,
+      "step": 21799
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8990681171417236,
+      "learning_rate": 4.4240021147237644e-05,
+      "loss": 1.3418,
+      "step": 21800
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7552669048309326,
+      "learning_rate": 4.423975680676712e-05,
+      "loss": 1.265,
+      "step": 21801
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6999386548995972,
+      "learning_rate": 4.42394924662966e-05,
+      "loss": 1.2123,
+      "step": 21802
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9285757541656494,
+      "learning_rate": 4.4239228125826065e-05,
+      "loss": 1.3182,
+      "step": 21803
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.891207218170166,
+      "learning_rate": 4.423896378535554e-05,
+      "loss": 2.234,
+      "step": 21804
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8437703847885132,
+      "learning_rate": 4.423869944488501e-05,
+      "loss": 2.1556,
+      "step": 21805
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1358957290649414,
+      "learning_rate": 4.4238435104414486e-05,
+      "loss": 1.383,
+      "step": 21806
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9892257452011108,
+      "learning_rate": 4.4238170763943966e-05,
+      "loss": 1.2671,
+      "step": 21807
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6433494091033936,
+      "learning_rate": 4.423790642347343e-05,
+      "loss": 1.1422,
+      "step": 21808
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.1430327892303467,
+      "learning_rate": 4.423764208300291e-05,
+      "loss": 1.3199,
+      "step": 21809
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.2613565921783447,
+      "learning_rate": 4.423737774253238e-05,
+      "loss": 1.4181,
+      "step": 21810
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.482792615890503,
+      "learning_rate": 4.423711340206186e-05,
+      "loss": 1.2961,
+      "step": 21811
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.829210042953491,
+      "learning_rate": 4.4236849061591334e-05,
+      "loss": 1.4943,
+      "step": 21812
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9395053386688232,
+      "learning_rate": 4.423658472112081e-05,
+      "loss": 1.3562,
+      "step": 21813
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5677096843719482,
+      "learning_rate": 4.423632038065028e-05,
+      "loss": 1.2033,
+      "step": 21814
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8615491390228271,
+      "learning_rate": 4.4236056040179754e-05,
+      "loss": 1.2372,
+      "step": 21815
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.712457299232483,
+      "learning_rate": 4.423579169970923e-05,
+      "loss": 1.1384,
+      "step": 21816
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4278602600097656,
+      "learning_rate": 4.42355273592387e-05,
+      "loss": 1.2954,
+      "step": 21817
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.099207878112793,
+      "learning_rate": 4.4235263018768175e-05,
+      "loss": 1.4371,
+      "step": 21818
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.3275671005249023,
+      "learning_rate": 4.423499867829765e-05,
+      "loss": 1.3419,
+      "step": 21819
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.277233600616455,
+      "learning_rate": 4.423473433782712e-05,
+      "loss": 1.3819,
+      "step": 21820
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.54618763923645,
+      "learning_rate": 4.4234469997356596e-05,
+      "loss": 1.3926,
+      "step": 21821
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8668501377105713,
+      "learning_rate": 4.4234205656886076e-05,
+      "loss": 1.2719,
+      "step": 21822
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8569819927215576,
+      "learning_rate": 4.423394131641554e-05,
+      "loss": 1.371,
+      "step": 21823
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.07853364944458,
+      "learning_rate": 4.423367697594502e-05,
+      "loss": 1.3091,
+      "step": 21824
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0021109580993652,
+      "learning_rate": 4.423341263547449e-05,
+      "loss": 1.4374,
+      "step": 21825
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.647266149520874,
+      "learning_rate": 4.423314829500397e-05,
+      "loss": 1.2827,
+      "step": 21826
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8148648738861084,
+      "learning_rate": 4.4232883954533444e-05,
+      "loss": 1.1843,
+      "step": 21827
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7448368072509766,
+      "learning_rate": 4.423261961406292e-05,
+      "loss": 1.2471,
+      "step": 21828
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.3674795627593994,
+      "learning_rate": 4.423235527359239e-05,
+      "loss": 1.417,
+      "step": 21829
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.14023756980896,
+      "learning_rate": 4.423209093312186e-05,
+      "loss": 1.3831,
+      "step": 21830
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.061211109161377,
+      "learning_rate": 4.423182659265134e-05,
+      "loss": 1.2104,
+      "step": 21831
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.162557363510132,
+      "learning_rate": 4.423156225218081e-05,
+      "loss": 1.3496,
+      "step": 21832
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6277399063110352,
+      "learning_rate": 4.4231297911710285e-05,
+      "loss": 1.0684,
+      "step": 21833
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.564765214920044,
+      "learning_rate": 4.423103357123976e-05,
+      "loss": 1.2844,
+      "step": 21834
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8345155715942383,
+      "learning_rate": 4.423076923076923e-05,
+      "loss": 1.3563,
+      "step": 21835
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8802077770233154,
+      "learning_rate": 4.4230504890298706e-05,
+      "loss": 1.3329,
+      "step": 21836
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.878432273864746,
+      "learning_rate": 4.4230240549828187e-05,
+      "loss": 1.2532,
+      "step": 21837
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5273845195770264,
+      "learning_rate": 4.422997620935765e-05,
+      "loss": 1.8627,
+      "step": 21838
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5099620819091797,
+      "learning_rate": 4.4229711868887134e-05,
+      "loss": 1.5041,
+      "step": 21839
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8331611156463623,
+      "learning_rate": 4.42294475284166e-05,
+      "loss": 1.3347,
+      "step": 21840
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8936374187469482,
+      "learning_rate": 4.4229183187946074e-05,
+      "loss": 1.2674,
+      "step": 21841
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8351184129714966,
+      "learning_rate": 4.4228918847475554e-05,
+      "loss": 1.1741,
+      "step": 21842
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7967462539672852,
+      "learning_rate": 4.422865450700502e-05,
+      "loss": 1.3805,
+      "step": 21843
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.832287073135376,
+      "learning_rate": 4.42283901665345e-05,
+      "loss": 1.194,
+      "step": 21844
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2747061252593994,
+      "learning_rate": 4.422812582606397e-05,
+      "loss": 1.3702,
+      "step": 21845
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0787553787231445,
+      "learning_rate": 4.422786148559345e-05,
+      "loss": 1.1917,
+      "step": 21846
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2975399494171143,
+      "learning_rate": 4.422759714512292e-05,
+      "loss": 1.4123,
+      "step": 21847
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.180077075958252,
+      "learning_rate": 4.4227332804652396e-05,
+      "loss": 1.3519,
+      "step": 21848
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8872339725494385,
+      "learning_rate": 4.422706846418187e-05,
+      "loss": 1.3219,
+      "step": 21849
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0259110927581787,
+      "learning_rate": 4.422680412371134e-05,
+      "loss": 1.3703,
+      "step": 21850
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6966692209243774,
+      "learning_rate": 4.4226539783240817e-05,
+      "loss": 1.4722,
+      "step": 21851
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.958940029144287,
+      "learning_rate": 4.422627544277029e-05,
+      "loss": 1.3506,
+      "step": 21852
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9883593320846558,
+      "learning_rate": 4.4226011102299764e-05,
+      "loss": 1.2275,
+      "step": 21853
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.062455654144287,
+      "learning_rate": 4.422574676182924e-05,
+      "loss": 1.3715,
+      "step": 21854
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.74612295627594,
+      "learning_rate": 4.422548242135871e-05,
+      "loss": 0.974,
+      "step": 21855
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0997610092163086,
+      "learning_rate": 4.4225218080888184e-05,
+      "loss": 1.4113,
+      "step": 21856
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.076284646987915,
+      "learning_rate": 4.4224953740417665e-05,
+      "loss": 1.2343,
+      "step": 21857
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.422905206680298,
+      "learning_rate": 4.422468939994713e-05,
+      "loss": 1.4796,
+      "step": 21858
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7036279439926147,
+      "learning_rate": 4.422442505947661e-05,
+      "loss": 1.2037,
+      "step": 21859
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0713770389556885,
+      "learning_rate": 4.422416071900608e-05,
+      "loss": 1.3581,
+      "step": 21860
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.563061237335205,
+      "learning_rate": 4.422389637853556e-05,
+      "loss": 1.0932,
+      "step": 21861
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.3439087867736816,
+      "learning_rate": 4.422363203806503e-05,
+      "loss": 1.508,
+      "step": 21862
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.005612373352051,
+      "learning_rate": 4.4223367697594506e-05,
+      "loss": 1.2851,
+      "step": 21863
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1971232891082764,
+      "learning_rate": 4.422310335712398e-05,
+      "loss": 1.3825,
+      "step": 21864
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.029538869857788,
+      "learning_rate": 4.422283901665345e-05,
+      "loss": 1.3645,
+      "step": 21865
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.320685863494873,
+      "learning_rate": 4.422257467618293e-05,
+      "loss": 1.2721,
+      "step": 21866
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.6231143474578857,
+      "learning_rate": 4.42223103357124e-05,
+      "loss": 1.288,
+      "step": 21867
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1328377723693848,
+      "learning_rate": 4.4222045995241874e-05,
+      "loss": 1.3731,
+      "step": 21868
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.8120784759521484,
+      "learning_rate": 4.422178165477135e-05,
+      "loss": 1.6216,
+      "step": 21869
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6357895135879517,
+      "learning_rate": 4.422151731430082e-05,
+      "loss": 1.162,
+      "step": 21870
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.181349515914917,
+      "learning_rate": 4.4221252973830295e-05,
+      "loss": 1.3483,
+      "step": 21871
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.427394390106201,
+      "learning_rate": 4.4220988633359775e-05,
+      "loss": 1.5282,
+      "step": 21872
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.252250909805298,
+      "learning_rate": 4.422072429288924e-05,
+      "loss": 1.4859,
+      "step": 21873
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.042271852493286,
+      "learning_rate": 4.422045995241872e-05,
+      "loss": 1.3608,
+      "step": 21874
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9091519117355347,
+      "learning_rate": 4.422019561194819e-05,
+      "loss": 1.2071,
+      "step": 21875
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.805577278137207,
+      "learning_rate": 4.421993127147766e-05,
+      "loss": 2.059,
+      "step": 21876
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.169207811355591,
+      "learning_rate": 4.421966693100714e-05,
+      "loss": 1.4364,
+      "step": 21877
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0030455589294434,
+      "learning_rate": 4.421940259053661e-05,
+      "loss": 1.1545,
+      "step": 21878
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.1572322845458984,
+      "learning_rate": 4.421913825006609e-05,
+      "loss": 1.4724,
+      "step": 21879
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.022949695587158,
+      "learning_rate": 4.421887390959556e-05,
+      "loss": 1.3683,
+      "step": 21880
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9019747972488403,
+      "learning_rate": 4.421860956912504e-05,
+      "loss": 1.2901,
+      "step": 21881
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.623713731765747,
+      "learning_rate": 4.4218345228654504e-05,
+      "loss": 1.3881,
+      "step": 21882
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.920617699623108,
+      "learning_rate": 4.4218080888183984e-05,
+      "loss": 1.2174,
+      "step": 21883
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7383174896240234,
+      "learning_rate": 4.421781654771346e-05,
+      "loss": 1.2192,
+      "step": 21884
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8480207920074463,
+      "learning_rate": 4.421755220724293e-05,
+      "loss": 1.1757,
+      "step": 21885
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.005666494369507,
+      "learning_rate": 4.4217287866772405e-05,
+      "loss": 1.2136,
+      "step": 21886
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6778695583343506,
+      "learning_rate": 4.421702352630188e-05,
+      "loss": 1.2892,
+      "step": 21887
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.4630101919174194,
+      "learning_rate": 4.421675918583135e-05,
+      "loss": 1.0807,
+      "step": 21888
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.981557369232178,
+      "learning_rate": 4.4216494845360826e-05,
+      "loss": 1.6509,
+      "step": 21889
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7143800258636475,
+      "learning_rate": 4.42162305048903e-05,
+      "loss": 1.258,
+      "step": 21890
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1768176555633545,
+      "learning_rate": 4.421596616441977e-05,
+      "loss": 1.3817,
+      "step": 21891
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7518208026885986,
+      "learning_rate": 4.4215701823949246e-05,
+      "loss": 1.2211,
+      "step": 21892
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.152468204498291,
+      "learning_rate": 4.421543748347872e-05,
+      "loss": 1.3484,
+      "step": 21893
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.431351900100708,
+      "learning_rate": 4.42151731430082e-05,
+      "loss": 1.3328,
+      "step": 21894
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5171616077423096,
+      "learning_rate": 4.421490880253767e-05,
+      "loss": 1.382,
+      "step": 21895
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2552242279052734,
+      "learning_rate": 4.421464446206715e-05,
+      "loss": 1.256,
+      "step": 21896
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9176218509674072,
+      "learning_rate": 4.4214380121596614e-05,
+      "loss": 1.3262,
+      "step": 21897
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.618221640586853,
+      "learning_rate": 4.4214115781126095e-05,
+      "loss": 1.1011,
+      "step": 21898
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9127259254455566,
+      "learning_rate": 4.421385144065557e-05,
+      "loss": 1.1873,
+      "step": 21899
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.988898515701294,
+      "learning_rate": 4.421358710018504e-05,
+      "loss": 1.4102,
+      "step": 21900
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.381124258041382,
+      "learning_rate": 4.4213322759714515e-05,
+      "loss": 1.5215,
+      "step": 21901
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.1676483154296875,
+      "learning_rate": 4.421305841924399e-05,
+      "loss": 1.4965,
+      "step": 21902
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2208425998687744,
+      "learning_rate": 4.421279407877346e-05,
+      "loss": 1.4393,
+      "step": 21903
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9147449731826782,
+      "learning_rate": 4.4212529738302936e-05,
+      "loss": 1.2035,
+      "step": 21904
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0333757400512695,
+      "learning_rate": 4.421226539783241e-05,
+      "loss": 1.445,
+      "step": 21905
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.843938946723938,
+      "learning_rate": 4.421200105736188e-05,
+      "loss": 1.2326,
+      "step": 21906
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.839557409286499,
+      "learning_rate": 4.421173671689136e-05,
+      "loss": 1.2253,
+      "step": 21907
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.436514377593994,
+      "learning_rate": 4.421147237642083e-05,
+      "loss": 1.2872,
+      "step": 21908
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2582194805145264,
+      "learning_rate": 4.421120803595031e-05,
+      "loss": 1.4791,
+      "step": 21909
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.6629974842071533,
+      "learning_rate": 4.421094369547978e-05,
+      "loss": 1.4134,
+      "step": 21910
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.628493070602417,
+      "learning_rate": 4.421067935500926e-05,
+      "loss": 1.2817,
+      "step": 21911
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.812888264656067,
+      "learning_rate": 4.4210415014538725e-05,
+      "loss": 1.3742,
+      "step": 21912
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.00144624710083,
+      "learning_rate": 4.42101506740682e-05,
+      "loss": 1.2757,
+      "step": 21913
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.8069441318511963,
+      "learning_rate": 4.420988633359768e-05,
+      "loss": 1.3238,
+      "step": 21914
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2019314765930176,
+      "learning_rate": 4.4209621993127145e-05,
+      "loss": 1.3799,
+      "step": 21915
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.789667844772339,
+      "learning_rate": 4.4209357652656626e-05,
+      "loss": 1.3773,
+      "step": 21916
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8859009742736816,
+      "learning_rate": 4.420909331218609e-05,
+      "loss": 1.2256,
+      "step": 21917
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5798256397247314,
+      "learning_rate": 4.420882897171557e-05,
+      "loss": 1.3497,
+      "step": 21918
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8327281475067139,
+      "learning_rate": 4.4208564631245046e-05,
+      "loss": 1.179,
+      "step": 21919
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.255117893218994,
+      "learning_rate": 4.420830029077452e-05,
+      "loss": 1.3312,
+      "step": 21920
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.156310558319092,
+      "learning_rate": 4.4208035950303994e-05,
+      "loss": 1.3118,
+      "step": 21921
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8724846839904785,
+      "learning_rate": 4.420777160983347e-05,
+      "loss": 1.2519,
+      "step": 21922
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8893747329711914,
+      "learning_rate": 4.420750726936294e-05,
+      "loss": 1.1598,
+      "step": 21923
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9137409925460815,
+      "learning_rate": 4.4207242928892414e-05,
+      "loss": 1.1963,
+      "step": 21924
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7842001914978027,
+      "learning_rate": 4.420697858842189e-05,
+      "loss": 1.2091,
+      "step": 21925
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.81610107421875,
+      "learning_rate": 4.420671424795136e-05,
+      "loss": 1.1395,
+      "step": 21926
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7692824602127075,
+      "learning_rate": 4.4206449907480835e-05,
+      "loss": 1.295,
+      "step": 21927
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.767858624458313,
+      "learning_rate": 4.420618556701031e-05,
+      "loss": 1.3,
+      "step": 21928
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8166574239730835,
+      "learning_rate": 4.420592122653979e-05,
+      "loss": 1.2566,
+      "step": 21929
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5631303787231445,
+      "learning_rate": 4.4205656886069256e-05,
+      "loss": 1.3696,
+      "step": 21930
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.054460048675537,
+      "learning_rate": 4.4205392545598736e-05,
+      "loss": 1.3176,
+      "step": 21931
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9952834844589233,
+      "learning_rate": 4.42051282051282e-05,
+      "loss": 1.2603,
+      "step": 21932
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9452821016311646,
+      "learning_rate": 4.420486386465768e-05,
+      "loss": 1.9878,
+      "step": 21933
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2414562702178955,
+      "learning_rate": 4.420459952418716e-05,
+      "loss": 1.3651,
+      "step": 21934
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9797110557556152,
+      "learning_rate": 4.420433518371663e-05,
+      "loss": 1.2533,
+      "step": 21935
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.7430005073547363,
+      "learning_rate": 4.4204070843246104e-05,
+      "loss": 2.2345,
+      "step": 21936
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0642616748809814,
+      "learning_rate": 4.420380650277558e-05,
+      "loss": 1.3181,
+      "step": 21937
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.6001579761505127,
+      "learning_rate": 4.420354216230505e-05,
+      "loss": 1.1934,
+      "step": 21938
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.5397638082504272,
+      "learning_rate": 4.4203277821834525e-05,
+      "loss": 1.2638,
+      "step": 21939
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1500742435455322,
+      "learning_rate": 4.4203013481364e-05,
+      "loss": 1.3642,
+      "step": 21940
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.138129234313965,
+      "learning_rate": 4.420274914089347e-05,
+      "loss": 1.6313,
+      "step": 21941
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.831383228302002,
+      "learning_rate": 4.4202484800422945e-05,
+      "loss": 1.3261,
+      "step": 21942
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0586555004119873,
+      "learning_rate": 4.420222045995242e-05,
+      "loss": 1.3466,
+      "step": 21943
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.7899720668792725,
+      "learning_rate": 4.42019561194819e-05,
+      "loss": 1.2786,
+      "step": 21944
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2876689434051514,
+      "learning_rate": 4.4201691779011366e-05,
+      "loss": 1.3064,
+      "step": 21945
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.831440806388855,
+      "learning_rate": 4.4201427438540846e-05,
+      "loss": 1.1347,
+      "step": 21946
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8970829248428345,
+      "learning_rate": 4.420116309807031e-05,
+      "loss": 1.335,
+      "step": 21947
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.341057062149048,
+      "learning_rate": 4.420089875759979e-05,
+      "loss": 1.1752,
+      "step": 21948
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8043056726455688,
+      "learning_rate": 4.420063441712927e-05,
+      "loss": 1.2795,
+      "step": 21949
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.078291177749634,
+      "learning_rate": 4.4200370076658734e-05,
+      "loss": 1.4791,
+      "step": 21950
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8694645166397095,
+      "learning_rate": 4.4200105736188214e-05,
+      "loss": 1.2109,
+      "step": 21951
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0051491260528564,
+      "learning_rate": 4.419984139571768e-05,
+      "loss": 1.3326,
+      "step": 21952
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1773667335510254,
+      "learning_rate": 4.419957705524716e-05,
+      "loss": 1.4192,
+      "step": 21953
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6726906299591064,
+      "learning_rate": 4.4199312714776635e-05,
+      "loss": 1.1147,
+      "step": 21954
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8061716556549072,
+      "learning_rate": 4.419904837430611e-05,
+      "loss": 1.3735,
+      "step": 21955
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.163395643234253,
+      "learning_rate": 4.419878403383558e-05,
+      "loss": 1.3575,
+      "step": 21956
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.058889389038086,
+      "learning_rate": 4.4198519693365056e-05,
+      "loss": 1.3706,
+      "step": 21957
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7868887186050415,
+      "learning_rate": 4.419825535289453e-05,
+      "loss": 1.8633,
+      "step": 21958
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.877909541130066,
+      "learning_rate": 4.4197991012424e-05,
+      "loss": 1.131,
+      "step": 21959
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.8386952877044678,
+      "learning_rate": 4.4197726671953476e-05,
+      "loss": 1.4169,
+      "step": 21960
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.034022569656372,
+      "learning_rate": 4.419746233148295e-05,
+      "loss": 1.3156,
+      "step": 21961
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6311371326446533,
+      "learning_rate": 4.4197197991012424e-05,
+      "loss": 1.0888,
+      "step": 21962
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.080580711364746,
+      "learning_rate": 4.41969336505419e-05,
+      "loss": 1.2787,
+      "step": 21963
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1766598224639893,
+      "learning_rate": 4.419666931007138e-05,
+      "loss": 1.2282,
+      "step": 21964
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.937954306602478,
+      "learning_rate": 4.4196404969600844e-05,
+      "loss": 1.2235,
+      "step": 21965
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.004242420196533,
+      "learning_rate": 4.4196140629130325e-05,
+      "loss": 1.2555,
+      "step": 21966
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.993159532546997,
+      "learning_rate": 4.419587628865979e-05,
+      "loss": 1.3003,
+      "step": 21967
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0242037773132324,
+      "learning_rate": 4.419561194818927e-05,
+      "loss": 1.2353,
+      "step": 21968
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5406334400177,
+      "learning_rate": 4.4195347607718745e-05,
+      "loss": 1.4603,
+      "step": 21969
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.898339033126831,
+      "learning_rate": 4.419508326724822e-05,
+      "loss": 1.2581,
+      "step": 21970
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.872931957244873,
+      "learning_rate": 4.419481892677769e-05,
+      "loss": 1.1706,
+      "step": 21971
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8271678686141968,
+      "learning_rate": 4.4194554586307166e-05,
+      "loss": 1.3203,
+      "step": 21972
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.6363236904144287,
+      "learning_rate": 4.419429024583664e-05,
+      "loss": 1.5143,
+      "step": 21973
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7277263402938843,
+      "learning_rate": 4.419402590536611e-05,
+      "loss": 1.1578,
+      "step": 21974
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.098578453063965,
+      "learning_rate": 4.419376156489559e-05,
+      "loss": 1.3181,
+      "step": 21975
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.960566759109497,
+      "learning_rate": 4.419349722442506e-05,
+      "loss": 1.329,
+      "step": 21976
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8893684148788452,
+      "learning_rate": 4.4193232883954534e-05,
+      "loss": 1.2552,
+      "step": 21977
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9535750150680542,
+      "learning_rate": 4.419296854348401e-05,
+      "loss": 1.2598,
+      "step": 21978
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6102550029754639,
+      "learning_rate": 4.419270420301349e-05,
+      "loss": 1.1544,
+      "step": 21979
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.3281941413879395,
+      "learning_rate": 4.4192439862542955e-05,
+      "loss": 1.3821,
+      "step": 21980
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1549108028411865,
+      "learning_rate": 4.4192175522072435e-05,
+      "loss": 2.1138,
+      "step": 21981
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.3324670791625977,
+      "learning_rate": 4.41919111816019e-05,
+      "loss": 1.3768,
+      "step": 21982
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6616777181625366,
+      "learning_rate": 4.419164684113138e-05,
+      "loss": 1.4501,
+      "step": 21983
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.3130624294281006,
+      "learning_rate": 4.4191382500660856e-05,
+      "loss": 1.2925,
+      "step": 21984
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.2441935539245605,
+      "learning_rate": 4.419111816019032e-05,
+      "loss": 1.3562,
+      "step": 21985
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0435361862182617,
+      "learning_rate": 4.41908538197198e-05,
+      "loss": 1.3522,
+      "step": 21986
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.300262689590454,
+      "learning_rate": 4.419058947924927e-05,
+      "loss": 1.3877,
+      "step": 21987
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0081026554107666,
+      "learning_rate": 4.419032513877875e-05,
+      "loss": 1.231,
+      "step": 21988
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.073824167251587,
+      "learning_rate": 4.4190060798308224e-05,
+      "loss": 1.3013,
+      "step": 21989
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8843932151794434,
+      "learning_rate": 4.41897964578377e-05,
+      "loss": 1.2697,
+      "step": 21990
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8065376281738281,
+      "learning_rate": 4.418953211736717e-05,
+      "loss": 1.1852,
+      "step": 21991
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.014512538909912,
+      "learning_rate": 4.4189267776896644e-05,
+      "loss": 1.2826,
+      "step": 21992
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.040717124938965,
+      "learning_rate": 4.418900343642612e-05,
+      "loss": 1.2358,
+      "step": 21993
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.5213878154754639,
+      "learning_rate": 4.418873909595559e-05,
+      "loss": 1.2499,
+      "step": 21994
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1350820064544678,
+      "learning_rate": 4.4188474755485065e-05,
+      "loss": 1.1607,
+      "step": 21995
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8881016969680786,
+      "learning_rate": 4.418821041501454e-05,
+      "loss": 1.375,
+      "step": 21996
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5371553897857666,
+      "learning_rate": 4.418794607454401e-05,
+      "loss": 1.3103,
+      "step": 21997
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9130332469940186,
+      "learning_rate": 4.4187681734073486e-05,
+      "loss": 1.1506,
+      "step": 21998
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9921847581863403,
+      "learning_rate": 4.4187417393602966e-05,
+      "loss": 1.3049,
+      "step": 21999
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.989190697669983,
+      "learning_rate": 4.418715305313243e-05,
+      "loss": 1.8673,
+      "step": 22000
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.393277883529663,
+      "learning_rate": 4.418688871266191e-05,
+      "loss": 1.2515,
+      "step": 22001
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.185818910598755,
+      "learning_rate": 4.418662437219138e-05,
+      "loss": 2.9716,
+      "step": 22002
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8928626775741577,
+      "learning_rate": 4.418636003172086e-05,
+      "loss": 1.2616,
+      "step": 22003
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.017595052719116,
+      "learning_rate": 4.4186095691250334e-05,
+      "loss": 1.3567,
+      "step": 22004
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2287228107452393,
+      "learning_rate": 4.418583135077981e-05,
+      "loss": 1.3647,
+      "step": 22005
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0559499263763428,
+      "learning_rate": 4.418556701030928e-05,
+      "loss": 1.147,
+      "step": 22006
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.771069884300232,
+      "learning_rate": 4.4185302669838755e-05,
+      "loss": 1.1101,
+      "step": 22007
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0245959758758545,
+      "learning_rate": 4.418503832936823e-05,
+      "loss": 1.3316,
+      "step": 22008
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.3535425662994385,
+      "learning_rate": 4.41847739888977e-05,
+      "loss": 1.4232,
+      "step": 22009
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8416272401809692,
+      "learning_rate": 4.4184509648427175e-05,
+      "loss": 1.2802,
+      "step": 22010
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.791920781135559,
+      "learning_rate": 4.418424530795665e-05,
+      "loss": 1.2721,
+      "step": 22011
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.157968759536743,
+      "learning_rate": 4.418398096748612e-05,
+      "loss": 1.2902,
+      "step": 22012
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.199753999710083,
+      "learning_rate": 4.4183716627015596e-05,
+      "loss": 2.6728,
+      "step": 22013
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7240267992019653,
+      "learning_rate": 4.4183452286545076e-05,
+      "loss": 1.1881,
+      "step": 22014
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.812367558479309,
+      "learning_rate": 4.418318794607454e-05,
+      "loss": 1.2439,
+      "step": 22015
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8583450317382812,
+      "learning_rate": 4.4182923605604024e-05,
+      "loss": 1.1917,
+      "step": 22016
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.951176404953003,
+      "learning_rate": 4.418265926513349e-05,
+      "loss": 1.286,
+      "step": 22017
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.03263783454895,
+      "learning_rate": 4.418239492466297e-05,
+      "loss": 1.3634,
+      "step": 22018
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.095165491104126,
+      "learning_rate": 4.4182130584192444e-05,
+      "loss": 1.4023,
+      "step": 22019
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.085768699645996,
+      "learning_rate": 4.418186624372192e-05,
+      "loss": 1.4158,
+      "step": 22020
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9179773330688477,
+      "learning_rate": 4.418160190325139e-05,
+      "loss": 1.2402,
+      "step": 22021
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.295673131942749,
+      "learning_rate": 4.418133756278086e-05,
+      "loss": 1.4435,
+      "step": 22022
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0771737098693848,
+      "learning_rate": 4.418107322231034e-05,
+      "loss": 1.4361,
+      "step": 22023
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0613081455230713,
+      "learning_rate": 4.418080888183981e-05,
+      "loss": 1.2066,
+      "step": 22024
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7633341550827026,
+      "learning_rate": 4.4180544541369286e-05,
+      "loss": 1.2153,
+      "step": 22025
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.958239197731018,
+      "learning_rate": 4.418028020089876e-05,
+      "loss": 1.2048,
+      "step": 22026
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8484795093536377,
+      "learning_rate": 4.418001586042823e-05,
+      "loss": 1.297,
+      "step": 22027
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.617695927619934,
+      "learning_rate": 4.4179751519957706e-05,
+      "loss": 1.4611,
+      "step": 22028
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.759816884994507,
+      "learning_rate": 4.417948717948719e-05,
+      "loss": 1.4542,
+      "step": 22029
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8049505949020386,
+      "learning_rate": 4.4179222839016653e-05,
+      "loss": 1.3813,
+      "step": 22030
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8738502264022827,
+      "learning_rate": 4.417895849854613e-05,
+      "loss": 1.4221,
+      "step": 22031
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.018416166305542,
+      "learning_rate": 4.41786941580756e-05,
+      "loss": 1.2238,
+      "step": 22032
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.184704065322876,
+      "learning_rate": 4.4178429817605074e-05,
+      "loss": 1.4745,
+      "step": 22033
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.296996593475342,
+      "learning_rate": 4.4178165477134555e-05,
+      "loss": 1.181,
+      "step": 22034
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0758519172668457,
+      "learning_rate": 4.417790113666402e-05,
+      "loss": 1.4611,
+      "step": 22035
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.830305337905884,
+      "learning_rate": 4.41776367961935e-05,
+      "loss": 1.3965,
+      "step": 22036
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.878203868865967,
+      "learning_rate": 4.417737245572297e-05,
+      "loss": 1.3611,
+      "step": 22037
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1285080909729004,
+      "learning_rate": 4.417710811525245e-05,
+      "loss": 1.2129,
+      "step": 22038
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9863696098327637,
+      "learning_rate": 4.417684377478192e-05,
+      "loss": 1.3482,
+      "step": 22039
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2540483474731445,
+      "learning_rate": 4.4176579434311396e-05,
+      "loss": 1.4213,
+      "step": 22040
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.150447368621826,
+      "learning_rate": 4.417631509384087e-05,
+      "loss": 1.3586,
+      "step": 22041
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.838047742843628,
+      "learning_rate": 4.417605075337034e-05,
+      "loss": 1.3655,
+      "step": 22042
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.981366753578186,
+      "learning_rate": 4.417578641289982e-05,
+      "loss": 1.403,
+      "step": 22043
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9692962169647217,
+      "learning_rate": 4.417552207242929e-05,
+      "loss": 1.2269,
+      "step": 22044
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.021437644958496,
+      "learning_rate": 4.4175257731958764e-05,
+      "loss": 1.2622,
+      "step": 22045
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8468117713928223,
+      "learning_rate": 4.417499339148824e-05,
+      "loss": 1.874,
+      "step": 22046
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1674482822418213,
+      "learning_rate": 4.417472905101771e-05,
+      "loss": 1.2586,
+      "step": 22047
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.142853021621704,
+      "learning_rate": 4.4174464710547185e-05,
+      "loss": 1.41,
+      "step": 22048
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.778432011604309,
+      "learning_rate": 4.4174200370076665e-05,
+      "loss": 1.2753,
+      "step": 22049
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2262396812438965,
+      "learning_rate": 4.417393602960613e-05,
+      "loss": 1.1518,
+      "step": 22050
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.49099862575531,
+      "learning_rate": 4.417367168913561e-05,
+      "loss": 1.1047,
+      "step": 22051
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8805410861968994,
+      "learning_rate": 4.417340734866508e-05,
+      "loss": 1.194,
+      "step": 22052
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9734050035476685,
+      "learning_rate": 4.417314300819456e-05,
+      "loss": 1.2769,
+      "step": 22053
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.277886390686035,
+      "learning_rate": 4.417287866772403e-05,
+      "loss": 1.5202,
+      "step": 22054
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.022740125656128,
+      "learning_rate": 4.4172614327253506e-05,
+      "loss": 2.3769,
+      "step": 22055
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0150036811828613,
+      "learning_rate": 4.417234998678298e-05,
+      "loss": 1.3053,
+      "step": 22056
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9187031984329224,
+      "learning_rate": 4.417208564631245e-05,
+      "loss": 1.3506,
+      "step": 22057
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.6287580728530884,
+      "learning_rate": 4.417182130584193e-05,
+      "loss": 1.243,
+      "step": 22058
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.1450178623199463,
+      "learning_rate": 4.41715569653714e-05,
+      "loss": 1.419,
+      "step": 22059
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7762458324432373,
+      "learning_rate": 4.4171292624900874e-05,
+      "loss": 1.1544,
+      "step": 22060
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.133805513381958,
+      "learning_rate": 4.417102828443035e-05,
+      "loss": 1.4062,
+      "step": 22061
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7983088493347168,
+      "learning_rate": 4.417076394395982e-05,
+      "loss": 1.3081,
+      "step": 22062
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0313620567321777,
+      "learning_rate": 4.4170499603489295e-05,
+      "loss": 1.1999,
+      "step": 22063
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9783341884613037,
+      "learning_rate": 4.4170235263018775e-05,
+      "loss": 1.3685,
+      "step": 22064
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1003506183624268,
+      "learning_rate": 4.416997092254824e-05,
+      "loss": 1.2627,
+      "step": 22065
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9840012788772583,
+      "learning_rate": 4.4169706582077716e-05,
+      "loss": 1.3475,
+      "step": 22066
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.3930108547210693,
+      "learning_rate": 4.416944224160719e-05,
+      "loss": 1.3731,
+      "step": 22067
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1276214122772217,
+      "learning_rate": 4.416917790113666e-05,
+      "loss": 1.3787,
+      "step": 22068
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1698873043060303,
+      "learning_rate": 4.416891356066614e-05,
+      "loss": 1.3124,
+      "step": 22069
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9290701150894165,
+      "learning_rate": 4.416864922019561e-05,
+      "loss": 1.2145,
+      "step": 22070
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0448014736175537,
+      "learning_rate": 4.416838487972509e-05,
+      "loss": 1.375,
+      "step": 22071
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.113511085510254,
+      "learning_rate": 4.416812053925456e-05,
+      "loss": 1.2293,
+      "step": 22072
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7861524820327759,
+      "learning_rate": 4.416785619878404e-05,
+      "loss": 1.1896,
+      "step": 22073
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.5785176753997803,
+      "learning_rate": 4.416759185831351e-05,
+      "loss": 1.6373,
+      "step": 22074
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2697300910949707,
+      "learning_rate": 4.4167327517842985e-05,
+      "loss": 1.3591,
+      "step": 22075
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.867696285247803,
+      "learning_rate": 4.416706317737246e-05,
+      "loss": 1.3891,
+      "step": 22076
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1140334606170654,
+      "learning_rate": 4.416679883690193e-05,
+      "loss": 1.3629,
+      "step": 22077
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.14878249168396,
+      "learning_rate": 4.4166534496431405e-05,
+      "loss": 1.776,
+      "step": 22078
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.20828914642334,
+      "learning_rate": 4.416627015596088e-05,
+      "loss": 1.1709,
+      "step": 22079
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0255954265594482,
+      "learning_rate": 4.416600581549035e-05,
+      "loss": 1.2716,
+      "step": 22080
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9026222229003906,
+      "learning_rate": 4.4165741475019826e-05,
+      "loss": 1.2909,
+      "step": 22081
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.903145670890808,
+      "learning_rate": 4.41654771345493e-05,
+      "loss": 1.276,
+      "step": 22082
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.321194648742676,
+      "learning_rate": 4.416521279407877e-05,
+      "loss": 1.4028,
+      "step": 22083
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0172553062438965,
+      "learning_rate": 4.4164948453608253e-05,
+      "loss": 1.245,
+      "step": 22084
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.5252618789672852,
+      "learning_rate": 4.416468411313772e-05,
+      "loss": 1.6661,
+      "step": 22085
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.9626471996307373,
+      "learning_rate": 4.41644197726672e-05,
+      "loss": 1.5564,
+      "step": 22086
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9872915744781494,
+      "learning_rate": 4.416415543219667e-05,
+      "loss": 1.4133,
+      "step": 22087
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.64599871635437,
+      "learning_rate": 4.416389109172615e-05,
+      "loss": 1.3829,
+      "step": 22088
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8955005407333374,
+      "learning_rate": 4.416362675125562e-05,
+      "loss": 1.3854,
+      "step": 22089
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7200536727905273,
+      "learning_rate": 4.4163362410785095e-05,
+      "loss": 1.1405,
+      "step": 22090
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.308368444442749,
+      "learning_rate": 4.416309807031457e-05,
+      "loss": 1.074,
+      "step": 22091
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0495922565460205,
+      "learning_rate": 4.416283372984404e-05,
+      "loss": 1.3123,
+      "step": 22092
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7702147960662842,
+      "learning_rate": 4.4162569389373516e-05,
+      "loss": 1.2,
+      "step": 22093
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.080228805541992,
+      "learning_rate": 4.416230504890299e-05,
+      "loss": 1.3147,
+      "step": 22094
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.970583200454712,
+      "learning_rate": 4.416204070843246e-05,
+      "loss": 1.1224,
+      "step": 22095
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.967160701751709,
+      "learning_rate": 4.4161776367961936e-05,
+      "loss": 1.3194,
+      "step": 22096
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0092427730560303,
+      "learning_rate": 4.416151202749141e-05,
+      "loss": 1.4397,
+      "step": 22097
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2893638610839844,
+      "learning_rate": 4.4161247687020883e-05,
+      "loss": 1.2147,
+      "step": 22098
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9831125736236572,
+      "learning_rate": 4.4160983346550364e-05,
+      "loss": 1.3077,
+      "step": 22099
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2638394832611084,
+      "learning_rate": 4.416071900607983e-05,
+      "loss": 1.3208,
+      "step": 22100
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.687963604927063,
+      "learning_rate": 4.416045466560931e-05,
+      "loss": 1.2605,
+      "step": 22101
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8998706340789795,
+      "learning_rate": 4.416019032513878e-05,
+      "loss": 1.0769,
+      "step": 22102
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0751705169677734,
+      "learning_rate": 4.415992598466825e-05,
+      "loss": 1.3495,
+      "step": 22103
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.808104395866394,
+      "learning_rate": 4.415966164419773e-05,
+      "loss": 2.0,
+      "step": 22104
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.819310188293457,
+      "learning_rate": 4.41593973037272e-05,
+      "loss": 1.3552,
+      "step": 22105
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9628645181655884,
+      "learning_rate": 4.415913296325668e-05,
+      "loss": 1.2882,
+      "step": 22106
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.157482862472534,
+      "learning_rate": 4.4158868622786146e-05,
+      "loss": 1.271,
+      "step": 22107
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9303820133209229,
+      "learning_rate": 4.4158604282315626e-05,
+      "loss": 1.2907,
+      "step": 22108
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.020902395248413,
+      "learning_rate": 4.41583399418451e-05,
+      "loss": 1.3706,
+      "step": 22109
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.4879672527313232,
+      "learning_rate": 4.415807560137457e-05,
+      "loss": 1.3783,
+      "step": 22110
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.191263198852539,
+      "learning_rate": 4.415781126090405e-05,
+      "loss": 1.1965,
+      "step": 22111
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1212306022644043,
+      "learning_rate": 4.415754692043352e-05,
+      "loss": 2.2133,
+      "step": 22112
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.5430532693862915,
+      "learning_rate": 4.4157282579962994e-05,
+      "loss": 1.2678,
+      "step": 22113
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1366052627563477,
+      "learning_rate": 4.415701823949247e-05,
+      "loss": 1.2564,
+      "step": 22114
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1935489177703857,
+      "learning_rate": 4.415675389902194e-05,
+      "loss": 1.2165,
+      "step": 22115
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.792036771774292,
+      "learning_rate": 4.4156489558551414e-05,
+      "loss": 1.2509,
+      "step": 22116
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.271017074584961,
+      "learning_rate": 4.415622521808089e-05,
+      "loss": 1.344,
+      "step": 22117
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.014507532119751,
+      "learning_rate": 4.415596087761036e-05,
+      "loss": 1.2318,
+      "step": 22118
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9397615194320679,
+      "learning_rate": 4.415569653713984e-05,
+      "loss": 1.4026,
+      "step": 22119
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9657354354858398,
+      "learning_rate": 4.415543219666931e-05,
+      "loss": 1.348,
+      "step": 22120
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0207574367523193,
+      "learning_rate": 4.415516785619879e-05,
+      "loss": 1.419,
+      "step": 22121
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.424959897994995,
+      "learning_rate": 4.4154903515728256e-05,
+      "loss": 1.3703,
+      "step": 22122
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.139404058456421,
+      "learning_rate": 4.4154639175257736e-05,
+      "loss": 1.3034,
+      "step": 22123
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.795514464378357,
+      "learning_rate": 4.415437483478721e-05,
+      "loss": 1.2305,
+      "step": 22124
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8739862442016602,
+      "learning_rate": 4.4154110494316683e-05,
+      "loss": 1.3191,
+      "step": 22125
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1239025592803955,
+      "learning_rate": 4.415384615384616e-05,
+      "loss": 1.3181,
+      "step": 22126
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.364492654800415,
+      "learning_rate": 4.415358181337563e-05,
+      "loss": 1.3213,
+      "step": 22127
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.493542432785034,
+      "learning_rate": 4.4153317472905104e-05,
+      "loss": 1.3582,
+      "step": 22128
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2425365447998047,
+      "learning_rate": 4.415305313243458e-05,
+      "loss": 1.2917,
+      "step": 22129
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.384467840194702,
+      "learning_rate": 4.415278879196405e-05,
+      "loss": 1.3732,
+      "step": 22130
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2157933712005615,
+      "learning_rate": 4.4152524451493525e-05,
+      "loss": 1.3144,
+      "step": 22131
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9077595472335815,
+      "learning_rate": 4.4152260111023e-05,
+      "loss": 1.2134,
+      "step": 22132
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1418230533599854,
+      "learning_rate": 4.415199577055247e-05,
+      "loss": 1.3558,
+      "step": 22133
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.322535991668701,
+      "learning_rate": 4.415173143008195e-05,
+      "loss": 1.298,
+      "step": 22134
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9930423498153687,
+      "learning_rate": 4.415146708961142e-05,
+      "loss": 1.3071,
+      "step": 22135
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.6692988872528076,
+      "learning_rate": 4.41512027491409e-05,
+      "loss": 1.6857,
+      "step": 22136
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.3718302249908447,
+      "learning_rate": 4.4150938408670366e-05,
+      "loss": 1.4446,
+      "step": 22137
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9637444019317627,
+      "learning_rate": 4.415067406819984e-05,
+      "loss": 1.3844,
+      "step": 22138
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0863325595855713,
+      "learning_rate": 4.415040972772932e-05,
+      "loss": 1.3225,
+      "step": 22139
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2215399742126465,
+      "learning_rate": 4.415014538725879e-05,
+      "loss": 2.0933,
+      "step": 22140
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.171332836151123,
+      "learning_rate": 4.414988104678827e-05,
+      "loss": 1.3286,
+      "step": 22141
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9632914066314697,
+      "learning_rate": 4.4149616706317734e-05,
+      "loss": 1.3077,
+      "step": 22142
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.332515001296997,
+      "learning_rate": 4.4149352365847214e-05,
+      "loss": 1.5234,
+      "step": 22143
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.088141441345215,
+      "learning_rate": 4.414908802537669e-05,
+      "loss": 1.3293,
+      "step": 22144
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.6677861213684082,
+      "learning_rate": 4.414882368490616e-05,
+      "loss": 1.1925,
+      "step": 22145
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.009739637374878,
+      "learning_rate": 4.4148559344435635e-05,
+      "loss": 1.3182,
+      "step": 22146
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.5893343687057495,
+      "learning_rate": 4.414829500396511e-05,
+      "loss": 1.6888,
+      "step": 22147
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.181248188018799,
+      "learning_rate": 4.414803066349458e-05,
+      "loss": 1.3133,
+      "step": 22148
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9721404314041138,
+      "learning_rate": 4.4147766323024056e-05,
+      "loss": 1.2597,
+      "step": 22149
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.73329758644104,
+      "learning_rate": 4.414750198255353e-05,
+      "loss": 1.2587,
+      "step": 22150
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8464444875717163,
+      "learning_rate": 4.4147237642083e-05,
+      "loss": 1.2905,
+      "step": 22151
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.051835536956787,
+      "learning_rate": 4.4146973301612477e-05,
+      "loss": 1.3318,
+      "step": 22152
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.779118537902832,
+      "learning_rate": 4.414670896114195e-05,
+      "loss": 1.34,
+      "step": 22153
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.398097038269043,
+      "learning_rate": 4.414644462067143e-05,
+      "loss": 1.3928,
+      "step": 22154
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1684253215789795,
+      "learning_rate": 4.41461802802009e-05,
+      "loss": 1.2565,
+      "step": 22155
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.079598903656006,
+      "learning_rate": 4.414591593973038e-05,
+      "loss": 1.4047,
+      "step": 22156
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1011877059936523,
+      "learning_rate": 4.4145651599259844e-05,
+      "loss": 1.4421,
+      "step": 22157
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1578245162963867,
+      "learning_rate": 4.4145387258789325e-05,
+      "loss": 1.204,
+      "step": 22158
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.801461935043335,
+      "learning_rate": 4.41451229183188e-05,
+      "loss": 1.6295,
+      "step": 22159
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.141960620880127,
+      "learning_rate": 4.414485857784827e-05,
+      "loss": 1.2626,
+      "step": 22160
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.007075071334839,
+      "learning_rate": 4.4144594237377745e-05,
+      "loss": 1.3167,
+      "step": 22161
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.977046251296997,
+      "learning_rate": 4.414432989690722e-05,
+      "loss": 1.1529,
+      "step": 22162
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.3694889545440674,
+      "learning_rate": 4.414406555643669e-05,
+      "loss": 1.367,
+      "step": 22163
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0917630195617676,
+      "learning_rate": 4.4143801215966166e-05,
+      "loss": 1.3419,
+      "step": 22164
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.748179316520691,
+      "learning_rate": 4.414353687549564e-05,
+      "loss": 1.3748,
+      "step": 22165
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9336990118026733,
+      "learning_rate": 4.414327253502511e-05,
+      "loss": 1.4001,
+      "step": 22166
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.335434913635254,
+      "learning_rate": 4.414300819455459e-05,
+      "loss": 1.2438,
+      "step": 22167
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.422551393508911,
+      "learning_rate": 4.414274385408406e-05,
+      "loss": 1.4045,
+      "step": 22168
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.6828105449676514,
+      "learning_rate": 4.414247951361354e-05,
+      "loss": 1.3824,
+      "step": 22169
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.355286121368408,
+      "learning_rate": 4.414221517314301e-05,
+      "loss": 1.3292,
+      "step": 22170
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.727593421936035,
+      "learning_rate": 4.414195083267249e-05,
+      "loss": 1.3075,
+      "step": 22171
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.036461591720581,
+      "learning_rate": 4.4141686492201955e-05,
+      "loss": 1.2634,
+      "step": 22172
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7811495065689087,
+      "learning_rate": 4.4141422151731435e-05,
+      "loss": 1.3489,
+      "step": 22173
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.00679087638855,
+      "learning_rate": 4.414115781126091e-05,
+      "loss": 1.2988,
+      "step": 22174
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.054473876953125,
+      "learning_rate": 4.4140893470790375e-05,
+      "loss": 1.2175,
+      "step": 22175
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.766993522644043,
+      "learning_rate": 4.4140629130319856e-05,
+      "loss": 1.1475,
+      "step": 22176
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.927342653274536,
+      "learning_rate": 4.414036478984932e-05,
+      "loss": 1.4584,
+      "step": 22177
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.435556411743164,
+      "learning_rate": 4.41401004493788e-05,
+      "loss": 1.3342,
+      "step": 22178
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.202364683151245,
+      "learning_rate": 4.4139836108908277e-05,
+      "loss": 1.5081,
+      "step": 22179
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.231125593185425,
+      "learning_rate": 4.413957176843775e-05,
+      "loss": 1.4259,
+      "step": 22180
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.25594162940979,
+      "learning_rate": 4.4139307427967224e-05,
+      "loss": 1.5336,
+      "step": 22181
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9333395957946777,
+      "learning_rate": 4.41390430874967e-05,
+      "loss": 1.3257,
+      "step": 22182
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.38985276222229,
+      "learning_rate": 4.413877874702617e-05,
+      "loss": 1.4167,
+      "step": 22183
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.089052200317383,
+      "learning_rate": 4.4138514406555644e-05,
+      "loss": 1.3516,
+      "step": 22184
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0847368240356445,
+      "learning_rate": 4.413825006608512e-05,
+      "loss": 1.3183,
+      "step": 22185
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.4798431396484375,
+      "learning_rate": 4.413798572561459e-05,
+      "loss": 1.2948,
+      "step": 22186
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7451571226119995,
+      "learning_rate": 4.4137721385144065e-05,
+      "loss": 1.1898,
+      "step": 22187
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.11700439453125,
+      "learning_rate": 4.413745704467354e-05,
+      "loss": 1.4394,
+      "step": 22188
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.4436874389648438,
+      "learning_rate": 4.413719270420302e-05,
+      "loss": 1.3624,
+      "step": 22189
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.294098377227783,
+      "learning_rate": 4.4136928363732486e-05,
+      "loss": 1.2581,
+      "step": 22190
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9170702695846558,
+      "learning_rate": 4.4136664023261966e-05,
+      "loss": 1.2886,
+      "step": 22191
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.866895079612732,
+      "learning_rate": 4.413639968279143e-05,
+      "loss": 1.2178,
+      "step": 22192
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.5106160640716553,
+      "learning_rate": 4.413613534232091e-05,
+      "loss": 1.1912,
+      "step": 22193
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.460805892944336,
+      "learning_rate": 4.413587100185039e-05,
+      "loss": 1.0835,
+      "step": 22194
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.6733797788619995,
+      "learning_rate": 4.413560666137986e-05,
+      "loss": 1.1342,
+      "step": 22195
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1159563064575195,
+      "learning_rate": 4.4135342320909334e-05,
+      "loss": 1.2688,
+      "step": 22196
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1407692432403564,
+      "learning_rate": 4.413507798043881e-05,
+      "loss": 1.3496,
+      "step": 22197
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2756009101867676,
+      "learning_rate": 4.413481363996828e-05,
+      "loss": 1.2951,
+      "step": 22198
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8537745475769043,
+      "learning_rate": 4.4134549299497755e-05,
+      "loss": 1.3014,
+      "step": 22199
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1324541568756104,
+      "learning_rate": 4.413428495902723e-05,
+      "loss": 1.4173,
+      "step": 22200
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8778635263442993,
+      "learning_rate": 4.41340206185567e-05,
+      "loss": 1.284,
+      "step": 22201
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1557838916778564,
+      "learning_rate": 4.4133756278086175e-05,
+      "loss": 1.4491,
+      "step": 22202
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2533762454986572,
+      "learning_rate": 4.413349193761565e-05,
+      "loss": 1.337,
+      "step": 22203
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8349884748458862,
+      "learning_rate": 4.413322759714513e-05,
+      "loss": 1.3545,
+      "step": 22204
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.275876998901367,
+      "learning_rate": 4.4132963256674596e-05,
+      "loss": 1.3452,
+      "step": 22205
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.404053211212158,
+      "learning_rate": 4.4132698916204077e-05,
+      "loss": 1.4557,
+      "step": 22206
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.5227229595184326,
+      "learning_rate": 4.413243457573354e-05,
+      "loss": 1.3573,
+      "step": 22207
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.408128023147583,
+      "learning_rate": 4.4132170235263024e-05,
+      "loss": 1.4059,
+      "step": 22208
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8469696044921875,
+      "learning_rate": 4.41319058947925e-05,
+      "loss": 1.4001,
+      "step": 22209
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1864445209503174,
+      "learning_rate": 4.413164155432197e-05,
+      "loss": 1.2751,
+      "step": 22210
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9792695045471191,
+      "learning_rate": 4.4131377213851444e-05,
+      "loss": 1.3726,
+      "step": 22211
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8648120164871216,
+      "learning_rate": 4.413111287338091e-05,
+      "loss": 1.2396,
+      "step": 22212
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.179527521133423,
+      "learning_rate": 4.413084853291039e-05,
+      "loss": 1.1334,
+      "step": 22213
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.26987886428833,
+      "learning_rate": 4.4130584192439865e-05,
+      "loss": 1.4033,
+      "step": 22214
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0522584915161133,
+      "learning_rate": 4.413031985196934e-05,
+      "loss": 1.2696,
+      "step": 22215
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.9315046072006226,
+      "learning_rate": 4.413005551149881e-05,
+      "loss": 1.2065,
+      "step": 22216
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.6583855152130127,
+      "learning_rate": 4.4129791171028286e-05,
+      "loss": 1.8626,
+      "step": 22217
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.3092293739318848,
+      "learning_rate": 4.412952683055776e-05,
+      "loss": 1.3297,
+      "step": 22218
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.325105905532837,
+      "learning_rate": 4.412926249008724e-05,
+      "loss": 1.815,
+      "step": 22219
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.071687936782837,
+      "learning_rate": 4.4128998149616707e-05,
+      "loss": 1.4146,
+      "step": 22220
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.873272180557251,
+      "learning_rate": 4.412873380914618e-05,
+      "loss": 1.3071,
+      "step": 22221
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8709189891815186,
+      "learning_rate": 4.4128469468675654e-05,
+      "loss": 1.2364,
+      "step": 22222
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1269686222076416,
+      "learning_rate": 4.412820512820513e-05,
+      "loss": 1.2332,
+      "step": 22223
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0758697986602783,
+      "learning_rate": 4.412794078773461e-05,
+      "loss": 1.4066,
+      "step": 22224
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1729609966278076,
+      "learning_rate": 4.4127676447264074e-05,
+      "loss": 1.2723,
+      "step": 22225
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8279078006744385,
+      "learning_rate": 4.4127412106793555e-05,
+      "loss": 1.2254,
+      "step": 22226
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.451449155807495,
+      "learning_rate": 4.412714776632302e-05,
+      "loss": 1.4502,
+      "step": 22227
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9983232021331787,
+      "learning_rate": 4.41268834258525e-05,
+      "loss": 1.4612,
+      "step": 22228
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.5735169649124146,
+      "learning_rate": 4.4126619085381975e-05,
+      "loss": 1.181,
+      "step": 22229
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.416956663131714,
+      "learning_rate": 4.412635474491145e-05,
+      "loss": 1.4261,
+      "step": 22230
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2371037006378174,
+      "learning_rate": 4.412609040444092e-05,
+      "loss": 1.3653,
+      "step": 22231
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8103759288787842,
+      "learning_rate": 4.4125826063970396e-05,
+      "loss": 1.1736,
+      "step": 22232
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.3217215538024902,
+      "learning_rate": 4.412556172349987e-05,
+      "loss": 1.413,
+      "step": 22233
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9693936109542847,
+      "learning_rate": 4.412529738302934e-05,
+      "loss": 1.253,
+      "step": 22234
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8754247426986694,
+      "learning_rate": 4.412503304255882e-05,
+      "loss": 1.1931,
+      "step": 22235
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.636059045791626,
+      "learning_rate": 4.412476870208829e-05,
+      "loss": 1.3541,
+      "step": 22236
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1015729904174805,
+      "learning_rate": 4.4124504361617764e-05,
+      "loss": 1.2876,
+      "step": 22237
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0805952548980713,
+      "learning_rate": 4.412424002114724e-05,
+      "loss": 1.2167,
+      "step": 22238
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8660739660263062,
+      "learning_rate": 4.412397568067672e-05,
+      "loss": 1.3383,
+      "step": 22239
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7081924676895142,
+      "learning_rate": 4.4123711340206185e-05,
+      "loss": 1.3131,
+      "step": 22240
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.365114688873291,
+      "learning_rate": 4.4123446999735665e-05,
+      "loss": 2.1471,
+      "step": 22241
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.5295348167419434,
+      "learning_rate": 4.412318265926513e-05,
+      "loss": 1.4177,
+      "step": 22242
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7093826532363892,
+      "learning_rate": 4.412291831879461e-05,
+      "loss": 1.1789,
+      "step": 22243
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.275315284729004,
+      "learning_rate": 4.4122653978324086e-05,
+      "loss": 1.6288,
+      "step": 22244
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1201868057250977,
+      "learning_rate": 4.412238963785356e-05,
+      "loss": 1.57,
+      "step": 22245
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0635578632354736,
+      "learning_rate": 4.412212529738303e-05,
+      "loss": 1.3911,
+      "step": 22246
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.788254976272583,
+      "learning_rate": 4.41218609569125e-05,
+      "loss": 1.2624,
+      "step": 22247
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6043660640716553,
+      "learning_rate": 4.412159661644198e-05,
+      "loss": 2.1709,
+      "step": 22248
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.4904289245605469,
+      "learning_rate": 4.4121332275971454e-05,
+      "loss": 1.162,
+      "step": 22249
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3321311473846436,
+      "learning_rate": 4.412106793550093e-05,
+      "loss": 1.2235,
+      "step": 22250
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8533250093460083,
+      "learning_rate": 4.41208035950304e-05,
+      "loss": 1.1986,
+      "step": 22251
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.521702766418457,
+      "learning_rate": 4.4120539254559874e-05,
+      "loss": 1.2748,
+      "step": 22252
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.249434232711792,
+      "learning_rate": 4.412027491408935e-05,
+      "loss": 1.239,
+      "step": 22253
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.5425302982330322,
+      "learning_rate": 4.412001057361883e-05,
+      "loss": 2.116,
+      "step": 22254
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1136891841888428,
+      "learning_rate": 4.4119746233148295e-05,
+      "loss": 1.4846,
+      "step": 22255
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.664287805557251,
+      "learning_rate": 4.411948189267777e-05,
+      "loss": 1.3386,
+      "step": 22256
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.338174819946289,
+      "learning_rate": 4.411921755220724e-05,
+      "loss": 1.4084,
+      "step": 22257
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8324676752090454,
+      "learning_rate": 4.4118953211736716e-05,
+      "loss": 1.2389,
+      "step": 22258
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9231899976730347,
+      "learning_rate": 4.4118688871266196e-05,
+      "loss": 1.2665,
+      "step": 22259
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.819193959236145,
+      "learning_rate": 4.411842453079566e-05,
+      "loss": 1.355,
+      "step": 22260
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3257858753204346,
+      "learning_rate": 4.411816019032514e-05,
+      "loss": 1.379,
+      "step": 22261
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.156433343887329,
+      "learning_rate": 4.411789584985461e-05,
+      "loss": 1.5808,
+      "step": 22262
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.5280489921569824,
+      "learning_rate": 4.411763150938409e-05,
+      "loss": 1.2069,
+      "step": 22263
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3910436630249023,
+      "learning_rate": 4.4117367168913564e-05,
+      "loss": 1.4509,
+      "step": 22264
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.072875738143921,
+      "learning_rate": 4.411710282844304e-05,
+      "loss": 1.2682,
+      "step": 22265
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.014259099960327,
+      "learning_rate": 4.411683848797251e-05,
+      "loss": 1.3952,
+      "step": 22266
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.780164122581482,
+      "learning_rate": 4.4116574147501985e-05,
+      "loss": 1.1862,
+      "step": 22267
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.309932231903076,
+      "learning_rate": 4.411630980703146e-05,
+      "loss": 1.4586,
+      "step": 22268
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8258328437805176,
+      "learning_rate": 4.411604546656093e-05,
+      "loss": 1.129,
+      "step": 22269
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8575958013534546,
+      "learning_rate": 4.4115781126090405e-05,
+      "loss": 1.2377,
+      "step": 22270
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.861764430999756,
+      "learning_rate": 4.411551678561988e-05,
+      "loss": 1.4325,
+      "step": 22271
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.745300054550171,
+      "learning_rate": 4.411525244514935e-05,
+      "loss": 1.9112,
+      "step": 22272
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.36006760597229,
+      "learning_rate": 4.4114988104678826e-05,
+      "loss": 1.3849,
+      "step": 22273
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9351330995559692,
+      "learning_rate": 4.4114723764208306e-05,
+      "loss": 1.3831,
+      "step": 22274
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8630719184875488,
+      "learning_rate": 4.411445942373777e-05,
+      "loss": 1.1946,
+      "step": 22275
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1522531509399414,
+      "learning_rate": 4.4114195083267254e-05,
+      "loss": 1.2581,
+      "step": 22276
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9424364566802979,
+      "learning_rate": 4.411393074279672e-05,
+      "loss": 1.3279,
+      "step": 22277
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9141700267791748,
+      "learning_rate": 4.41136664023262e-05,
+      "loss": 1.353,
+      "step": 22278
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.245760917663574,
+      "learning_rate": 4.4113402061855674e-05,
+      "loss": 1.2281,
+      "step": 22279
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.334022045135498,
+      "learning_rate": 4.411313772138515e-05,
+      "loss": 1.4352,
+      "step": 22280
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1107418537139893,
+      "learning_rate": 4.411287338091462e-05,
+      "loss": 1.2884,
+      "step": 22281
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.263986587524414,
+      "learning_rate": 4.4112609040444095e-05,
+      "loss": 1.4258,
+      "step": 22282
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.520660400390625,
+      "learning_rate": 4.411234469997357e-05,
+      "loss": 1.4898,
+      "step": 22283
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7834173440933228,
+      "learning_rate": 4.411208035950304e-05,
+      "loss": 1.2507,
+      "step": 22284
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3667800426483154,
+      "learning_rate": 4.4111816019032516e-05,
+      "loss": 1.2802,
+      "step": 22285
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1236417293548584,
+      "learning_rate": 4.411155167856199e-05,
+      "loss": 1.2459,
+      "step": 22286
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.998740792274475,
+      "learning_rate": 4.411128733809146e-05,
+      "loss": 1.285,
+      "step": 22287
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.000142812728882,
+      "learning_rate": 4.4111022997620936e-05,
+      "loss": 1.5225,
+      "step": 22288
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8628907203674316,
+      "learning_rate": 4.411075865715042e-05,
+      "loss": 1.3174,
+      "step": 22289
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8912237882614136,
+      "learning_rate": 4.4110494316679884e-05,
+      "loss": 1.3684,
+      "step": 22290
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.447429895401001,
+      "learning_rate": 4.4110229976209364e-05,
+      "loss": 1.3459,
+      "step": 22291
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3839755058288574,
+      "learning_rate": 4.410996563573883e-05,
+      "loss": 1.4325,
+      "step": 22292
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7580151557922363,
+      "learning_rate": 4.4109701295268304e-05,
+      "loss": 1.189,
+      "step": 22293
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8967581987380981,
+      "learning_rate": 4.4109436954797785e-05,
+      "loss": 1.15,
+      "step": 22294
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.860962152481079,
+      "learning_rate": 4.410917261432725e-05,
+      "loss": 1.2719,
+      "step": 22295
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9457621574401855,
+      "learning_rate": 4.410890827385673e-05,
+      "loss": 1.8435,
+      "step": 22296
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9008290767669678,
+      "learning_rate": 4.41086439333862e-05,
+      "loss": 1.3109,
+      "step": 22297
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1403815746307373,
+      "learning_rate": 4.410837959291568e-05,
+      "loss": 1.4139,
+      "step": 22298
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8790245056152344,
+      "learning_rate": 4.410811525244515e-05,
+      "loss": 1.2869,
+      "step": 22299
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1839890480041504,
+      "learning_rate": 4.4107850911974626e-05,
+      "loss": 1.2661,
+      "step": 22300
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.679756999015808,
+      "learning_rate": 4.41075865715041e-05,
+      "loss": 1.1774,
+      "step": 22301
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.657163143157959,
+      "learning_rate": 4.410732223103357e-05,
+      "loss": 1.2595,
+      "step": 22302
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0714876651763916,
+      "learning_rate": 4.410705789056305e-05,
+      "loss": 1.3031,
+      "step": 22303
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.815568685531616,
+      "learning_rate": 4.410679355009252e-05,
+      "loss": 1.4023,
+      "step": 22304
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9872573614120483,
+      "learning_rate": 4.4106529209621994e-05,
+      "loss": 1.7644,
+      "step": 22305
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.589080572128296,
+      "learning_rate": 4.410626486915147e-05,
+      "loss": 1.376,
+      "step": 22306
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1040616035461426,
+      "learning_rate": 4.410600052868094e-05,
+      "loss": 1.3354,
+      "step": 22307
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8232965469360352,
+      "learning_rate": 4.4105736188210415e-05,
+      "loss": 1.18,
+      "step": 22308
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.809935450553894,
+      "learning_rate": 4.4105471847739895e-05,
+      "loss": 1.3269,
+      "step": 22309
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.286121129989624,
+      "learning_rate": 4.410520750726936e-05,
+      "loss": 1.2127,
+      "step": 22310
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.254725217819214,
+      "learning_rate": 4.410494316679884e-05,
+      "loss": 1.423,
+      "step": 22311
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.7157912254333496,
+      "learning_rate": 4.410467882632831e-05,
+      "loss": 1.5053,
+      "step": 22312
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3587167263031006,
+      "learning_rate": 4.410441448585779e-05,
+      "loss": 1.2898,
+      "step": 22313
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1191587448120117,
+      "learning_rate": 4.410415014538726e-05,
+      "loss": 1.2414,
+      "step": 22314
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8448851108551025,
+      "learning_rate": 4.4103885804916736e-05,
+      "loss": 1.1612,
+      "step": 22315
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8857526779174805,
+      "learning_rate": 4.410362146444621e-05,
+      "loss": 1.2275,
+      "step": 22316
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2280800342559814,
+      "learning_rate": 4.4103357123975684e-05,
+      "loss": 1.2396,
+      "step": 22317
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8187123537063599,
+      "learning_rate": 4.410309278350516e-05,
+      "loss": 1.307,
+      "step": 22318
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0702505111694336,
+      "learning_rate": 4.410282844303463e-05,
+      "loss": 1.3747,
+      "step": 22319
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.9800262451171875,
+      "learning_rate": 4.4102564102564104e-05,
+      "loss": 1.2443,
+      "step": 22320
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2435882091522217,
+      "learning_rate": 4.410229976209358e-05,
+      "loss": 1.2567,
+      "step": 22321
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.089888572692871,
+      "learning_rate": 4.410203542162305e-05,
+      "loss": 1.2568,
+      "step": 22322
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7445502281188965,
+      "learning_rate": 4.4101771081152525e-05,
+      "loss": 1.2173,
+      "step": 22323
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.439631700515747,
+      "learning_rate": 4.4101506740682005e-05,
+      "loss": 1.4598,
+      "step": 22324
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9635157585144043,
+      "learning_rate": 4.410124240021147e-05,
+      "loss": 1.1634,
+      "step": 22325
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.268993854522705,
+      "learning_rate": 4.410097805974095e-05,
+      "loss": 1.4074,
+      "step": 22326
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9346907138824463,
+      "learning_rate": 4.410071371927042e-05,
+      "loss": 1.2447,
+      "step": 22327
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.15317440032959,
+      "learning_rate": 4.41004493787999e-05,
+      "loss": 1.235,
+      "step": 22328
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9723553657531738,
+      "learning_rate": 4.410018503832937e-05,
+      "loss": 1.3082,
+      "step": 22329
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9465233087539673,
+      "learning_rate": 4.409992069785884e-05,
+      "loss": 1.3411,
+      "step": 22330
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.5596234798431396,
+      "learning_rate": 4.409965635738832e-05,
+      "loss": 1.3602,
+      "step": 22331
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1951303482055664,
+      "learning_rate": 4.409939201691779e-05,
+      "loss": 1.3505,
+      "step": 22332
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.296205997467041,
+      "learning_rate": 4.409912767644727e-05,
+      "loss": 1.3116,
+      "step": 22333
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1526196002960205,
+      "learning_rate": 4.409886333597674e-05,
+      "loss": 1.3422,
+      "step": 22334
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.161327362060547,
+      "learning_rate": 4.4098598995506215e-05,
+      "loss": 1.1991,
+      "step": 22335
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0368239879608154,
+      "learning_rate": 4.409833465503569e-05,
+      "loss": 1.293,
+      "step": 22336
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.2647130489349365,
+      "learning_rate": 4.409807031456516e-05,
+      "loss": 1.4557,
+      "step": 22337
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.5721933841705322,
+      "learning_rate": 4.4097805974094635e-05,
+      "loss": 1.3669,
+      "step": 22338
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.014376163482666,
+      "learning_rate": 4.409754163362411e-05,
+      "loss": 1.3534,
+      "step": 22339
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3346309661865234,
+      "learning_rate": 4.409727729315358e-05,
+      "loss": 1.3273,
+      "step": 22340
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9245620965957642,
+      "learning_rate": 4.4097012952683056e-05,
+      "loss": 1.3613,
+      "step": 22341
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1062004566192627,
+      "learning_rate": 4.409674861221253e-05,
+      "loss": 1.3017,
+      "step": 22342
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.264610528945923,
+      "learning_rate": 4.4096484271742e-05,
+      "loss": 1.5021,
+      "step": 22343
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.9138288497924805,
+      "learning_rate": 4.4096219931271484e-05,
+      "loss": 1.4471,
+      "step": 22344
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9005869626998901,
+      "learning_rate": 4.409595559080095e-05,
+      "loss": 1.3122,
+      "step": 22345
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9541199207305908,
+      "learning_rate": 4.409569125033043e-05,
+      "loss": 1.2721,
+      "step": 22346
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1817805767059326,
+      "learning_rate": 4.40954269098599e-05,
+      "loss": 1.4079,
+      "step": 22347
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.976798415184021,
+      "learning_rate": 4.409516256938938e-05,
+      "loss": 1.3648,
+      "step": 22348
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6898980140686035,
+      "learning_rate": 4.409489822891885e-05,
+      "loss": 1.2844,
+      "step": 22349
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7220872640609741,
+      "learning_rate": 4.4094633888448325e-05,
+      "loss": 1.2526,
+      "step": 22350
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2352235317230225,
+      "learning_rate": 4.40943695479778e-05,
+      "loss": 1.2782,
+      "step": 22351
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.131300687789917,
+      "learning_rate": 4.409410520750727e-05,
+      "loss": 1.4854,
+      "step": 22352
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3392271995544434,
+      "learning_rate": 4.4093840867036746e-05,
+      "loss": 1.3736,
+      "step": 22353
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.774164915084839,
+      "learning_rate": 4.409357652656622e-05,
+      "loss": 1.5083,
+      "step": 22354
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.917873501777649,
+      "learning_rate": 4.409331218609569e-05,
+      "loss": 1.2535,
+      "step": 22355
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.900043249130249,
+      "learning_rate": 4.4093047845625166e-05,
+      "loss": 1.3713,
+      "step": 22356
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2888591289520264,
+      "learning_rate": 4.409278350515464e-05,
+      "loss": 1.4634,
+      "step": 22357
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.5959339141845703,
+      "learning_rate": 4.4092519164684114e-05,
+      "loss": 1.3737,
+      "step": 22358
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3327882289886475,
+      "learning_rate": 4.4092254824213594e-05,
+      "loss": 1.7478,
+      "step": 22359
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.304539680480957,
+      "learning_rate": 4.409199048374306e-05,
+      "loss": 1.3345,
+      "step": 22360
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.736665964126587,
+      "learning_rate": 4.409172614327254e-05,
+      "loss": 1.3862,
+      "step": 22361
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.7101070880889893,
+      "learning_rate": 4.409146180280201e-05,
+      "loss": 1.3991,
+      "step": 22362
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7107661962509155,
+      "learning_rate": 4.409119746233149e-05,
+      "loss": 1.29,
+      "step": 22363
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9292725324630737,
+      "learning_rate": 4.409093312186096e-05,
+      "loss": 1.2719,
+      "step": 22364
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.311405658721924,
+      "learning_rate": 4.409066878139043e-05,
+      "loss": 1.4733,
+      "step": 22365
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8032341003417969,
+      "learning_rate": 4.409040444091991e-05,
+      "loss": 1.2603,
+      "step": 22366
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.4188530445098877,
+      "learning_rate": 4.4090140100449376e-05,
+      "loss": 1.2825,
+      "step": 22367
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8577817678451538,
+      "learning_rate": 4.4089875759978856e-05,
+      "loss": 1.2435,
+      "step": 22368
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2803008556365967,
+      "learning_rate": 4.408961141950833e-05,
+      "loss": 1.4208,
+      "step": 22369
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0111422538757324,
+      "learning_rate": 4.40893470790378e-05,
+      "loss": 1.4073,
+      "step": 22370
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0028555393218994,
+      "learning_rate": 4.408908273856728e-05,
+      "loss": 1.281,
+      "step": 22371
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9887834787368774,
+      "learning_rate": 4.408881839809675e-05,
+      "loss": 1.2566,
+      "step": 22372
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.037435531616211,
+      "learning_rate": 4.4088554057626224e-05,
+      "loss": 1.317,
+      "step": 22373
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3545608520507812,
+      "learning_rate": 4.40882897171557e-05,
+      "loss": 1.2638,
+      "step": 22374
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1425864696502686,
+      "learning_rate": 4.408802537668517e-05,
+      "loss": 1.2431,
+      "step": 22375
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.249802589416504,
+      "learning_rate": 4.4087761036214645e-05,
+      "loss": 1.2517,
+      "step": 22376
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9961472749710083,
+      "learning_rate": 4.408749669574412e-05,
+      "loss": 1.1914,
+      "step": 22377
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9829626083374023,
+      "learning_rate": 4.408723235527359e-05,
+      "loss": 1.4648,
+      "step": 22378
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3400442600250244,
+      "learning_rate": 4.408696801480307e-05,
+      "loss": 1.4138,
+      "step": 22379
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.23661208152771,
+      "learning_rate": 4.408670367433254e-05,
+      "loss": 1.172,
+      "step": 22380
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7445694208145142,
+      "learning_rate": 4.408643933386202e-05,
+      "loss": 1.4495,
+      "step": 22381
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.228933572769165,
+      "learning_rate": 4.4086174993391486e-05,
+      "loss": 1.4285,
+      "step": 22382
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9928312301635742,
+      "learning_rate": 4.4085910652920966e-05,
+      "loss": 1.2684,
+      "step": 22383
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9668134450912476,
+      "learning_rate": 4.408564631245044e-05,
+      "loss": 2.1838,
+      "step": 22384
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8450201749801636,
+      "learning_rate": 4.4085381971979913e-05,
+      "loss": 1.3153,
+      "step": 22385
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.711594820022583,
+      "learning_rate": 4.408511763150939e-05,
+      "loss": 1.306,
+      "step": 22386
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7012319564819336,
+      "learning_rate": 4.408485329103886e-05,
+      "loss": 1.254,
+      "step": 22387
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.167332410812378,
+      "learning_rate": 4.4084588950568334e-05,
+      "loss": 1.2189,
+      "step": 22388
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.542675733566284,
+      "learning_rate": 4.408432461009781e-05,
+      "loss": 1.3516,
+      "step": 22389
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6992075443267822,
+      "learning_rate": 4.408406026962728e-05,
+      "loss": 1.2137,
+      "step": 22390
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0460867881774902,
+      "learning_rate": 4.4083795929156755e-05,
+      "loss": 1.2484,
+      "step": 22391
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9203221797943115,
+      "learning_rate": 4.408353158868623e-05,
+      "loss": 1.2332,
+      "step": 22392
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.714756965637207,
+      "learning_rate": 4.40832672482157e-05,
+      "loss": 1.3156,
+      "step": 22393
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.186115026473999,
+      "learning_rate": 4.408300290774518e-05,
+      "loss": 1.2524,
+      "step": 22394
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8193819522857666,
+      "learning_rate": 4.408273856727465e-05,
+      "loss": 1.0822,
+      "step": 22395
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1565191745758057,
+      "learning_rate": 4.408247422680413e-05,
+      "loss": 1.1924,
+      "step": 22396
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.5488897562026978,
+      "learning_rate": 4.4082209886333596e-05,
+      "loss": 1.149,
+      "step": 22397
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8510351181030273,
+      "learning_rate": 4.408194554586308e-05,
+      "loss": 1.2188,
+      "step": 22398
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.804368019104004,
+      "learning_rate": 4.408168120539255e-05,
+      "loss": 1.298,
+      "step": 22399
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9307469129562378,
+      "learning_rate": 4.4081416864922024e-05,
+      "loss": 1.2664,
+      "step": 22400
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6065120697021484,
+      "learning_rate": 4.40811525244515e-05,
+      "loss": 1.3213,
+      "step": 22401
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0152204036712646,
+      "learning_rate": 4.4080888183980964e-05,
+      "loss": 1.1881,
+      "step": 22402
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.823110818862915,
+      "learning_rate": 4.4080623843510445e-05,
+      "loss": 1.2452,
+      "step": 22403
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8553931713104248,
+      "learning_rate": 4.408035950303992e-05,
+      "loss": 1.2866,
+      "step": 22404
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.1352157592773438,
+      "learning_rate": 4.408009516256939e-05,
+      "loss": 1.3895,
+      "step": 22405
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.544487476348877,
+      "learning_rate": 4.4079830822098865e-05,
+      "loss": 1.1991,
+      "step": 22406
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.043809652328491,
+      "learning_rate": 4.407956648162834e-05,
+      "loss": 1.2388,
+      "step": 22407
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3963818550109863,
+      "learning_rate": 4.407930214115781e-05,
+      "loss": 1.3571,
+      "step": 22408
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9610713720321655,
+      "learning_rate": 4.407903780068729e-05,
+      "loss": 1.37,
+      "step": 22409
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8757823705673218,
+      "learning_rate": 4.407877346021676e-05,
+      "loss": 1.2717,
+      "step": 22410
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6694835424423218,
+      "learning_rate": 4.407850911974623e-05,
+      "loss": 1.1236,
+      "step": 22411
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.680417537689209,
+      "learning_rate": 4.407824477927571e-05,
+      "loss": 1.0687,
+      "step": 22412
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.587589740753174,
+      "learning_rate": 4.407798043880518e-05,
+      "loss": 1.3133,
+      "step": 22413
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8173047304153442,
+      "learning_rate": 4.407771609833466e-05,
+      "loss": 1.1702,
+      "step": 22414
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8532310724258423,
+      "learning_rate": 4.407745175786413e-05,
+      "loss": 1.2042,
+      "step": 22415
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0303149223327637,
+      "learning_rate": 4.407718741739361e-05,
+      "loss": 1.2959,
+      "step": 22416
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.6255242824554443,
+      "learning_rate": 4.4076923076923075e-05,
+      "loss": 0.9866,
+      "step": 22417
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.240586757659912,
+      "learning_rate": 4.4076658736452555e-05,
+      "loss": 1.3631,
+      "step": 22418
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.109807252883911,
+      "learning_rate": 4.407639439598203e-05,
+      "loss": 1.3531,
+      "step": 22419
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.494295597076416,
+      "learning_rate": 4.40761300555115e-05,
+      "loss": 1.2505,
+      "step": 22420
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.876434326171875,
+      "learning_rate": 4.4075865715040976e-05,
+      "loss": 1.2615,
+      "step": 22421
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.866881012916565,
+      "learning_rate": 4.407560137457045e-05,
+      "loss": 1.3321,
+      "step": 22422
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1688218116760254,
+      "learning_rate": 4.407533703409992e-05,
+      "loss": 1.2635,
+      "step": 22423
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9232633113861084,
+      "learning_rate": 4.4075072693629396e-05,
+      "loss": 1.3063,
+      "step": 22424
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.181248664855957,
+      "learning_rate": 4.407480835315887e-05,
+      "loss": 1.2799,
+      "step": 22425
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0712497234344482,
+      "learning_rate": 4.4074544012688343e-05,
+      "loss": 1.3544,
+      "step": 22426
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2354114055633545,
+      "learning_rate": 4.407427967221782e-05,
+      "loss": 1.2152,
+      "step": 22427
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.070250988006592,
+      "learning_rate": 4.407401533174729e-05,
+      "loss": 1.3152,
+      "step": 22428
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2050528526306152,
+      "learning_rate": 4.407375099127677e-05,
+      "loss": 1.2913,
+      "step": 22429
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3006725311279297,
+      "learning_rate": 4.407348665080624e-05,
+      "loss": 1.3943,
+      "step": 22430
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2265563011169434,
+      "learning_rate": 4.407322231033572e-05,
+      "loss": 1.2837,
+      "step": 22431
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.7300703525543213,
+      "learning_rate": 4.4072957969865185e-05,
+      "loss": 1.4099,
+      "step": 22432
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.079866409301758,
+      "learning_rate": 4.4072693629394665e-05,
+      "loss": 1.2483,
+      "step": 22433
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7819018363952637,
+      "learning_rate": 4.407242928892414e-05,
+      "loss": 1.358,
+      "step": 22434
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.172651529312134,
+      "learning_rate": 4.407216494845361e-05,
+      "loss": 1.2878,
+      "step": 22435
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8937718868255615,
+      "learning_rate": 4.4071900607983086e-05,
+      "loss": 1.153,
+      "step": 22436
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8328006267547607,
+      "learning_rate": 4.407163626751255e-05,
+      "loss": 1.2256,
+      "step": 22437
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3368701934814453,
+      "learning_rate": 4.407137192704203e-05,
+      "loss": 1.2972,
+      "step": 22438
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.6874337196350098,
+      "learning_rate": 4.407110758657151e-05,
+      "loss": 1.3191,
+      "step": 22439
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8754472732543945,
+      "learning_rate": 4.407084324610098e-05,
+      "loss": 1.2343,
+      "step": 22440
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1378071308135986,
+      "learning_rate": 4.4070578905630454e-05,
+      "loss": 1.2936,
+      "step": 22441
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1239593029022217,
+      "learning_rate": 4.407031456515993e-05,
+      "loss": 1.2421,
+      "step": 22442
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7011922597885132,
+      "learning_rate": 4.40700502246894e-05,
+      "loss": 1.1849,
+      "step": 22443
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0072860717773438,
+      "learning_rate": 4.406978588421888e-05,
+      "loss": 1.2423,
+      "step": 22444
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0266544818878174,
+      "learning_rate": 4.406952154374835e-05,
+      "loss": 1.3945,
+      "step": 22445
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4692816734313965,
+      "learning_rate": 4.406925720327783e-05,
+      "loss": 1.6787,
+      "step": 22446
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3420677185058594,
+      "learning_rate": 4.4068992862807295e-05,
+      "loss": 1.2013,
+      "step": 22447
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.207334518432617,
+      "learning_rate": 4.406872852233677e-05,
+      "loss": 1.1847,
+      "step": 22448
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7460765838623047,
+      "learning_rate": 4.406846418186625e-05,
+      "loss": 1.2312,
+      "step": 22449
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2989470958709717,
+      "learning_rate": 4.4068199841395716e-05,
+      "loss": 1.4637,
+      "step": 22450
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.089418411254883,
+      "learning_rate": 4.4067935500925196e-05,
+      "loss": 1.2654,
+      "step": 22451
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8745988607406616,
+      "learning_rate": 4.406767116045466e-05,
+      "loss": 1.3057,
+      "step": 22452
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.286867618560791,
+      "learning_rate": 4.4067406819984143e-05,
+      "loss": 1.4181,
+      "step": 22453
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.538618326187134,
+      "learning_rate": 4.406714247951362e-05,
+      "loss": 1.4036,
+      "step": 22454
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8635574579238892,
+      "learning_rate": 4.406687813904309e-05,
+      "loss": 1.3489,
+      "step": 22455
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7502188682556152,
+      "learning_rate": 4.4066613798572564e-05,
+      "loss": 1.7881,
+      "step": 22456
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1556854248046875,
+      "learning_rate": 4.406634945810204e-05,
+      "loss": 1.2789,
+      "step": 22457
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4387474060058594,
+      "learning_rate": 4.406608511763151e-05,
+      "loss": 1.2864,
+      "step": 22458
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0328428745269775,
+      "learning_rate": 4.4065820777160985e-05,
+      "loss": 1.3936,
+      "step": 22459
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.096233367919922,
+      "learning_rate": 4.406555643669046e-05,
+      "loss": 1.3348,
+      "step": 22460
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.6153342723846436,
+      "learning_rate": 4.406529209621993e-05,
+      "loss": 1.4274,
+      "step": 22461
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9700194597244263,
+      "learning_rate": 4.4065027755749406e-05,
+      "loss": 1.1795,
+      "step": 22462
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9296554327011108,
+      "learning_rate": 4.406476341527888e-05,
+      "loss": 1.2475,
+      "step": 22463
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.5580761432647705,
+      "learning_rate": 4.406449907480836e-05,
+      "loss": 1.0755,
+      "step": 22464
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.2476766109466553,
+      "learning_rate": 4.4064234734337826e-05,
+      "loss": 1.3779,
+      "step": 22465
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.984302043914795,
+      "learning_rate": 4.4063970393867307e-05,
+      "loss": 1.3198,
+      "step": 22466
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9833359718322754,
+      "learning_rate": 4.4063706053396773e-05,
+      "loss": 1.4963,
+      "step": 22467
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6845678091049194,
+      "learning_rate": 4.4063441712926254e-05,
+      "loss": 1.2941,
+      "step": 22468
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3710479736328125,
+      "learning_rate": 4.406317737245573e-05,
+      "loss": 1.3336,
+      "step": 22469
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6372382640838623,
+      "learning_rate": 4.40629130319852e-05,
+      "loss": 1.2124,
+      "step": 22470
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9770280122756958,
+      "learning_rate": 4.4062648691514674e-05,
+      "loss": 1.2143,
+      "step": 22471
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.986083745956421,
+      "learning_rate": 4.406238435104415e-05,
+      "loss": 1.3686,
+      "step": 22472
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8924864530563354,
+      "learning_rate": 4.406212001057362e-05,
+      "loss": 1.3707,
+      "step": 22473
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.605069875717163,
+      "learning_rate": 4.4061855670103095e-05,
+      "loss": 1.2536,
+      "step": 22474
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.756226658821106,
+      "learning_rate": 4.406159132963257e-05,
+      "loss": 1.2689,
+      "step": 22475
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7955689430236816,
+      "learning_rate": 4.406132698916204e-05,
+      "loss": 1.1807,
+      "step": 22476
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7762080430984497,
+      "learning_rate": 4.4061062648691516e-05,
+      "loss": 1.5006,
+      "step": 22477
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0713160037994385,
+      "learning_rate": 4.406079830822099e-05,
+      "loss": 1.4449,
+      "step": 22478
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9915790557861328,
+      "learning_rate": 4.406053396775047e-05,
+      "loss": 1.2788,
+      "step": 22479
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.051295042037964,
+      "learning_rate": 4.4060269627279937e-05,
+      "loss": 1.268,
+      "step": 22480
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.680541753768921,
+      "learning_rate": 4.406000528680942e-05,
+      "loss": 1.2162,
+      "step": 22481
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.607704758644104,
+      "learning_rate": 4.4059740946338884e-05,
+      "loss": 1.1454,
+      "step": 22482
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.695923924446106,
+      "learning_rate": 4.405947660586836e-05,
+      "loss": 1.4213,
+      "step": 22483
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8067491054534912,
+      "learning_rate": 4.405921226539784e-05,
+      "loss": 1.0975,
+      "step": 22484
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8109291791915894,
+      "learning_rate": 4.4058947924927304e-05,
+      "loss": 1.3108,
+      "step": 22485
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9884036779403687,
+      "learning_rate": 4.4058683584456785e-05,
+      "loss": 2.3911,
+      "step": 22486
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9132858514785767,
+      "learning_rate": 4.405841924398625e-05,
+      "loss": 1.1917,
+      "step": 22487
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4492218494415283,
+      "learning_rate": 4.405815490351573e-05,
+      "loss": 1.3752,
+      "step": 22488
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4965319633483887,
+      "learning_rate": 4.4057890563045206e-05,
+      "loss": 1.4311,
+      "step": 22489
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0517446994781494,
+      "learning_rate": 4.405762622257468e-05,
+      "loss": 1.3565,
+      "step": 22490
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.576704263687134,
+      "learning_rate": 4.405736188210415e-05,
+      "loss": 1.3389,
+      "step": 22491
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2840800285339355,
+      "learning_rate": 4.4057097541633626e-05,
+      "loss": 1.3641,
+      "step": 22492
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7406262159347534,
+      "learning_rate": 4.40568332011631e-05,
+      "loss": 1.1228,
+      "step": 22493
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9606469869613647,
+      "learning_rate": 4.405656886069257e-05,
+      "loss": 1.179,
+      "step": 22494
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.772596836090088,
+      "learning_rate": 4.405630452022205e-05,
+      "loss": 1.2156,
+      "step": 22495
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0595037937164307,
+      "learning_rate": 4.405604017975152e-05,
+      "loss": 1.3345,
+      "step": 22496
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.740445137023926,
+      "learning_rate": 4.4055775839280994e-05,
+      "loss": 1.1874,
+      "step": 22497
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.245957851409912,
+      "learning_rate": 4.405551149881047e-05,
+      "loss": 1.2986,
+      "step": 22498
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6447927951812744,
+      "learning_rate": 4.405524715833995e-05,
+      "loss": 1.3477,
+      "step": 22499
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.05410099029541,
+      "learning_rate": 4.4054982817869415e-05,
+      "loss": 1.2286,
+      "step": 22500
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3537559509277344,
+      "learning_rate": 4.4054718477398895e-05,
+      "loss": 1.409,
+      "step": 22501
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8145537376403809,
+      "learning_rate": 4.405445413692836e-05,
+      "loss": 1.3679,
+      "step": 22502
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9073952436447144,
+      "learning_rate": 4.405418979645784e-05,
+      "loss": 1.316,
+      "step": 22503
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3988423347473145,
+      "learning_rate": 4.4053925455987316e-05,
+      "loss": 2.0704,
+      "step": 22504
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9180200099945068,
+      "learning_rate": 4.405366111551679e-05,
+      "loss": 1.2226,
+      "step": 22505
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3904998302459717,
+      "learning_rate": 4.405339677504626e-05,
+      "loss": 1.3102,
+      "step": 22506
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8808051347732544,
+      "learning_rate": 4.4053132434575737e-05,
+      "loss": 1.3937,
+      "step": 22507
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0853824615478516,
+      "learning_rate": 4.405286809410521e-05,
+      "loss": 1.3372,
+      "step": 22508
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9565277099609375,
+      "learning_rate": 4.4052603753634684e-05,
+      "loss": 1.2977,
+      "step": 22509
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.090909004211426,
+      "learning_rate": 4.405233941316416e-05,
+      "loss": 1.234,
+      "step": 22510
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6282293796539307,
+      "learning_rate": 4.405207507269363e-05,
+      "loss": 1.1734,
+      "step": 22511
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9589189291000366,
+      "learning_rate": 4.4051810732223104e-05,
+      "loss": 1.2485,
+      "step": 22512
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4961671829223633,
+      "learning_rate": 4.405154639175258e-05,
+      "loss": 1.4269,
+      "step": 22513
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.285149574279785,
+      "learning_rate": 4.405128205128206e-05,
+      "loss": 1.3711,
+      "step": 22514
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9935287237167358,
+      "learning_rate": 4.4051017710811525e-05,
+      "loss": 1.3549,
+      "step": 22515
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0495550632476807,
+      "learning_rate": 4.4050753370341005e-05,
+      "loss": 1.3005,
+      "step": 22516
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.3816628456115723,
+      "learning_rate": 4.405048902987047e-05,
+      "loss": 1.5063,
+      "step": 22517
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.142735481262207,
+      "learning_rate": 4.405022468939995e-05,
+      "loss": 1.2709,
+      "step": 22518
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.683150053024292,
+      "learning_rate": 4.4049960348929426e-05,
+      "loss": 1.3526,
+      "step": 22519
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1318986415863037,
+      "learning_rate": 4.404969600845889e-05,
+      "loss": 1.3154,
+      "step": 22520
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.903791904449463,
+      "learning_rate": 4.404943166798837e-05,
+      "loss": 1.2518,
+      "step": 22521
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8064485788345337,
+      "learning_rate": 4.404916732751784e-05,
+      "loss": 1.294,
+      "step": 22522
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2450380325317383,
+      "learning_rate": 4.404890298704732e-05,
+      "loss": 1.4925,
+      "step": 22523
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3889195919036865,
+      "learning_rate": 4.4048638646576794e-05,
+      "loss": 1.3762,
+      "step": 22524
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.5749154090881348,
+      "learning_rate": 4.404837430610627e-05,
+      "loss": 1.1852,
+      "step": 22525
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7029873132705688,
+      "learning_rate": 4.404810996563574e-05,
+      "loss": 1.228,
+      "step": 22526
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8311405181884766,
+      "learning_rate": 4.4047845625165215e-05,
+      "loss": 1.3315,
+      "step": 22527
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.7879936695098877,
+      "learning_rate": 4.404758128469469e-05,
+      "loss": 1.2882,
+      "step": 22528
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.912736177444458,
+      "learning_rate": 4.404731694422416e-05,
+      "loss": 1.307,
+      "step": 22529
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2579195499420166,
+      "learning_rate": 4.4047052603753635e-05,
+      "loss": 1.3678,
+      "step": 22530
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.23842191696167,
+      "learning_rate": 4.404678826328311e-05,
+      "loss": 1.4091,
+      "step": 22531
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.370464324951172,
+      "learning_rate": 4.404652392281258e-05,
+      "loss": 1.2756,
+      "step": 22532
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1952731609344482,
+      "learning_rate": 4.4046259582342056e-05,
+      "loss": 1.2955,
+      "step": 22533
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6118911504745483,
+      "learning_rate": 4.4045995241871537e-05,
+      "loss": 1.2366,
+      "step": 22534
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9656349420547485,
+      "learning_rate": 4.4045730901401e-05,
+      "loss": 1.6592,
+      "step": 22535
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1325173377990723,
+      "learning_rate": 4.4045466560930484e-05,
+      "loss": 1.4869,
+      "step": 22536
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.666170597076416,
+      "learning_rate": 4.404520222045995e-05,
+      "loss": 1.3877,
+      "step": 22537
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.0220909118652344,
+      "learning_rate": 4.404493787998943e-05,
+      "loss": 1.3934,
+      "step": 22538
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9444741010665894,
+      "learning_rate": 4.4044673539518904e-05,
+      "loss": 2.2039,
+      "step": 22539
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.361978769302368,
+      "learning_rate": 4.404440919904838e-05,
+      "loss": 1.5582,
+      "step": 22540
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8283145427703857,
+      "learning_rate": 4.404414485857785e-05,
+      "loss": 1.3772,
+      "step": 22541
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4223358631134033,
+      "learning_rate": 4.4043880518107325e-05,
+      "loss": 1.3312,
+      "step": 22542
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.6931207180023193,
+      "learning_rate": 4.40436161776368e-05,
+      "loss": 1.3098,
+      "step": 22543
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9406511783599854,
+      "learning_rate": 4.404335183716627e-05,
+      "loss": 1.2957,
+      "step": 22544
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.225074529647827,
+      "learning_rate": 4.4043087496695746e-05,
+      "loss": 1.383,
+      "step": 22545
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.060838460922241,
+      "learning_rate": 4.404282315622522e-05,
+      "loss": 1.3529,
+      "step": 22546
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9712507724761963,
+      "learning_rate": 4.404255881575469e-05,
+      "loss": 1.262,
+      "step": 22547
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9688637256622314,
+      "learning_rate": 4.4042294475284167e-05,
+      "loss": 1.3087,
+      "step": 22548
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9486823081970215,
+      "learning_rate": 4.404203013481365e-05,
+      "loss": 1.3895,
+      "step": 22549
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.101132869720459,
+      "learning_rate": 4.4041765794343114e-05,
+      "loss": 1.1873,
+      "step": 22550
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.247067451477051,
+      "learning_rate": 4.4041501453872594e-05,
+      "loss": 1.3914,
+      "step": 22551
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6715418100357056,
+      "learning_rate": 4.404123711340206e-05,
+      "loss": 1.2162,
+      "step": 22552
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9638432264328003,
+      "learning_rate": 4.404097277293154e-05,
+      "loss": 1.2199,
+      "step": 22553
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.8330607414245605,
+      "learning_rate": 4.4040708432461015e-05,
+      "loss": 1.4697,
+      "step": 22554
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8174712657928467,
+      "learning_rate": 4.404044409199048e-05,
+      "loss": 1.4413,
+      "step": 22555
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0266482830047607,
+      "learning_rate": 4.404017975151996e-05,
+      "loss": 1.3424,
+      "step": 22556
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6818983554840088,
+      "learning_rate": 4.403991541104943e-05,
+      "loss": 1.3875,
+      "step": 22557
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3811228275299072,
+      "learning_rate": 4.403965107057891e-05,
+      "loss": 1.2738,
+      "step": 22558
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8178786039352417,
+      "learning_rate": 4.403938673010838e-05,
+      "loss": 1.2685,
+      "step": 22559
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8015681505203247,
+      "learning_rate": 4.4039122389637856e-05,
+      "loss": 1.3651,
+      "step": 22560
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.720395088195801,
+      "learning_rate": 4.403885804916733e-05,
+      "loss": 1.4925,
+      "step": 22561
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7785896062850952,
+      "learning_rate": 4.40385937086968e-05,
+      "loss": 1.1813,
+      "step": 22562
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0495753288269043,
+      "learning_rate": 4.403832936822628e-05,
+      "loss": 1.2368,
+      "step": 22563
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9141008853912354,
+      "learning_rate": 4.403806502775576e-05,
+      "loss": 1.4766,
+      "step": 22564
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.188577175140381,
+      "learning_rate": 4.4037800687285224e-05,
+      "loss": 1.2819,
+      "step": 22565
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7639129161834717,
+      "learning_rate": 4.40375363468147e-05,
+      "loss": 1.2543,
+      "step": 22566
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.7464025020599365,
+      "learning_rate": 4.403727200634417e-05,
+      "loss": 1.3467,
+      "step": 22567
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.798148274421692,
+      "learning_rate": 4.4037007665873645e-05,
+      "loss": 1.2638,
+      "step": 22568
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6948518753051758,
+      "learning_rate": 4.4036743325403125e-05,
+      "loss": 1.1907,
+      "step": 22569
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7994775772094727,
+      "learning_rate": 4.403647898493259e-05,
+      "loss": 1.2549,
+      "step": 22570
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.723648190498352,
+      "learning_rate": 4.403621464446207e-05,
+      "loss": 1.1951,
+      "step": 22571
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.913736343383789,
+      "learning_rate": 4.403595030399154e-05,
+      "loss": 1.3166,
+      "step": 22572
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.645780086517334,
+      "learning_rate": 4.403568596352102e-05,
+      "loss": 1.1752,
+      "step": 22573
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8851181268692017,
+      "learning_rate": 4.403542162305049e-05,
+      "loss": 1.2708,
+      "step": 22574
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.0685760974884033,
+      "learning_rate": 4.4035157282579966e-05,
+      "loss": 1.5754,
+      "step": 22575
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7272907495498657,
+      "learning_rate": 4.403489294210944e-05,
+      "loss": 1.1894,
+      "step": 22576
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1587345600128174,
+      "learning_rate": 4.4034628601638914e-05,
+      "loss": 1.3781,
+      "step": 22577
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.869185209274292,
+      "learning_rate": 4.403436426116839e-05,
+      "loss": 1.2599,
+      "step": 22578
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7647485733032227,
+      "learning_rate": 4.403409992069786e-05,
+      "loss": 1.2934,
+      "step": 22579
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.102003812789917,
+      "learning_rate": 4.4033835580227334e-05,
+      "loss": 1.3549,
+      "step": 22580
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8933169841766357,
+      "learning_rate": 4.403357123975681e-05,
+      "loss": 1.3269,
+      "step": 22581
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.6392228603363037,
+      "learning_rate": 4.403330689928628e-05,
+      "loss": 1.4019,
+      "step": 22582
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.085620641708374,
+      "learning_rate": 4.4033042558815755e-05,
+      "loss": 1.3966,
+      "step": 22583
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1056723594665527,
+      "learning_rate": 4.4032778218345235e-05,
+      "loss": 1.3761,
+      "step": 22584
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.209146738052368,
+      "learning_rate": 4.40325138778747e-05,
+      "loss": 1.2362,
+      "step": 22585
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.985642433166504,
+      "learning_rate": 4.403224953740418e-05,
+      "loss": 1.2741,
+      "step": 22586
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3223941326141357,
+      "learning_rate": 4.403198519693365e-05,
+      "loss": 1.2203,
+      "step": 22587
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2310235500335693,
+      "learning_rate": 4.403172085646313e-05,
+      "loss": 1.5323,
+      "step": 22588
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1724612712860107,
+      "learning_rate": 4.40314565159926e-05,
+      "loss": 1.4233,
+      "step": 22589
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.7171084880828857,
+      "learning_rate": 4.403119217552208e-05,
+      "loss": 1.5209,
+      "step": 22590
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0897557735443115,
+      "learning_rate": 4.403092783505155e-05,
+      "loss": 1.1885,
+      "step": 22591
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0721869468688965,
+      "learning_rate": 4.403066349458102e-05,
+      "loss": 1.3328,
+      "step": 22592
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9832327365875244,
+      "learning_rate": 4.40303991541105e-05,
+      "loss": 1.4013,
+      "step": 22593
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8978055715560913,
+      "learning_rate": 4.403013481363997e-05,
+      "loss": 1.268,
+      "step": 22594
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3263845443725586,
+      "learning_rate": 4.4029870473169445e-05,
+      "loss": 1.4259,
+      "step": 22595
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2117044925689697,
+      "learning_rate": 4.402960613269892e-05,
+      "loss": 1.2508,
+      "step": 22596
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4240434169769287,
+      "learning_rate": 4.402934179222839e-05,
+      "loss": 1.6306,
+      "step": 22597
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1371874809265137,
+      "learning_rate": 4.4029077451757865e-05,
+      "loss": 1.3063,
+      "step": 22598
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.988815188407898,
+      "learning_rate": 4.4028813111287346e-05,
+      "loss": 1.3752,
+      "step": 22599
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7100528478622437,
+      "learning_rate": 4.402854877081681e-05,
+      "loss": 1.1686,
+      "step": 22600
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.0623583793640137,
+      "learning_rate": 4.4028284430346286e-05,
+      "loss": 1.2003,
+      "step": 22601
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.3902175426483154,
+      "learning_rate": 4.402802008987576e-05,
+      "loss": 1.4001,
+      "step": 22602
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4847073554992676,
+      "learning_rate": 4.402775574940523e-05,
+      "loss": 1.4491,
+      "step": 22603
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.074859619140625,
+      "learning_rate": 4.4027491408934714e-05,
+      "loss": 1.3257,
+      "step": 22604
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.1805317401885986,
+      "learning_rate": 4.402722706846418e-05,
+      "loss": 1.3794,
+      "step": 22605
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8094027042388916,
+      "learning_rate": 4.402696272799366e-05,
+      "loss": 1.3064,
+      "step": 22606
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4153778553009033,
+      "learning_rate": 4.402669838752313e-05,
+      "loss": 1.2993,
+      "step": 22607
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8898319005966187,
+      "learning_rate": 4.402643404705261e-05,
+      "loss": 1.2801,
+      "step": 22608
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.234774351119995,
+      "learning_rate": 4.402616970658208e-05,
+      "loss": 1.361,
+      "step": 22609
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.409433126449585,
+      "learning_rate": 4.4025905366111555e-05,
+      "loss": 1.3997,
+      "step": 22610
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.9090750217437744,
+      "learning_rate": 4.402564102564103e-05,
+      "loss": 1.2608,
+      "step": 22611
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.760338544845581,
+      "learning_rate": 4.40253766851705e-05,
+      "loss": 1.0471,
+      "step": 22612
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.462681293487549,
+      "learning_rate": 4.4025112344699976e-05,
+      "loss": 1.3324,
+      "step": 22613
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.373997688293457,
+      "learning_rate": 4.402484800422945e-05,
+      "loss": 1.4199,
+      "step": 22614
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.711910605430603,
+      "learning_rate": 4.402458366375892e-05,
+      "loss": 1.2221,
+      "step": 22615
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.605219841003418,
+      "learning_rate": 4.4024319323288396e-05,
+      "loss": 1.3553,
+      "step": 22616
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.21331524848938,
+      "learning_rate": 4.402405498281787e-05,
+      "loss": 1.4588,
+      "step": 22617
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8006103038787842,
+      "learning_rate": 4.4023790642347344e-05,
+      "loss": 1.1921,
+      "step": 22618
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.886342167854309,
+      "learning_rate": 4.4023526301876824e-05,
+      "loss": 1.2412,
+      "step": 22619
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2750275135040283,
+      "learning_rate": 4.402326196140629e-05,
+      "loss": 1.3321,
+      "step": 22620
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7906156778335571,
+      "learning_rate": 4.402299762093577e-05,
+      "loss": 1.1351,
+      "step": 22621
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7944726943969727,
+      "learning_rate": 4.402273328046524e-05,
+      "loss": 1.2554,
+      "step": 22622
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1985697746276855,
+      "learning_rate": 4.402246893999472e-05,
+      "loss": 1.3617,
+      "step": 22623
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3159005641937256,
+      "learning_rate": 4.402220459952419e-05,
+      "loss": 1.8283,
+      "step": 22624
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8026005029678345,
+      "learning_rate": 4.4021940259053665e-05,
+      "loss": 1.3588,
+      "step": 22625
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8006620407104492,
+      "learning_rate": 4.402167591858314e-05,
+      "loss": 1.2003,
+      "step": 22626
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9034050703048706,
+      "learning_rate": 4.4021411578112606e-05,
+      "loss": 1.4886,
+      "step": 22627
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.073838233947754,
+      "learning_rate": 4.4021147237642086e-05,
+      "loss": 1.1911,
+      "step": 22628
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.9065229892730713,
+      "learning_rate": 4.402088289717156e-05,
+      "loss": 1.3213,
+      "step": 22629
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8477078676223755,
+      "learning_rate": 4.402061855670103e-05,
+      "loss": 1.2665,
+      "step": 22630
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.0095174312591553,
+      "learning_rate": 4.402035421623051e-05,
+      "loss": 1.3882,
+      "step": 22631
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9524606466293335,
+      "learning_rate": 4.402008987575998e-05,
+      "loss": 1.3297,
+      "step": 22632
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2418580055236816,
+      "learning_rate": 4.4019825535289454e-05,
+      "loss": 1.3901,
+      "step": 22633
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8027100563049316,
+      "learning_rate": 4.4019561194818934e-05,
+      "loss": 1.2467,
+      "step": 22634
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.892899990081787,
+      "learning_rate": 4.40192968543484e-05,
+      "loss": 1.2868,
+      "step": 22635
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.078096389770508,
+      "learning_rate": 4.401903251387788e-05,
+      "loss": 1.3004,
+      "step": 22636
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0290701389312744,
+      "learning_rate": 4.401876817340735e-05,
+      "loss": 1.2316,
+      "step": 22637
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.778357744216919,
+      "learning_rate": 4.401850383293682e-05,
+      "loss": 1.1457,
+      "step": 22638
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1675500869750977,
+      "learning_rate": 4.40182394924663e-05,
+      "loss": 1.3874,
+      "step": 22639
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3064119815826416,
+      "learning_rate": 4.401797515199577e-05,
+      "loss": 1.3587,
+      "step": 22640
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4830098152160645,
+      "learning_rate": 4.401771081152525e-05,
+      "loss": 1.3469,
+      "step": 22641
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.033597946166992,
+      "learning_rate": 4.4017446471054716e-05,
+      "loss": 1.4209,
+      "step": 22642
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2237298488616943,
+      "learning_rate": 4.4017182130584196e-05,
+      "loss": 1.3537,
+      "step": 22643
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1901352405548096,
+      "learning_rate": 4.401691779011367e-05,
+      "loss": 1.2577,
+      "step": 22644
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.155963659286499,
+      "learning_rate": 4.4016653449643144e-05,
+      "loss": 1.3193,
+      "step": 22645
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.94430673122406,
+      "learning_rate": 4.401638910917262e-05,
+      "loss": 1.3599,
+      "step": 22646
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.060345411300659,
+      "learning_rate": 4.401612476870209e-05,
+      "loss": 1.2466,
+      "step": 22647
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6096771955490112,
+      "learning_rate": 4.4015860428231564e-05,
+      "loss": 1.8754,
+      "step": 22648
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7255419492721558,
+      "learning_rate": 4.401559608776104e-05,
+      "loss": 1.2579,
+      "step": 22649
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7951915264129639,
+      "learning_rate": 4.401533174729051e-05,
+      "loss": 1.4055,
+      "step": 22650
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6919916868209839,
+      "learning_rate": 4.4015067406819985e-05,
+      "loss": 1.2652,
+      "step": 22651
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8366777896881104,
+      "learning_rate": 4.401480306634946e-05,
+      "loss": 1.2045,
+      "step": 22652
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9675421714782715,
+      "learning_rate": 4.401453872587893e-05,
+      "loss": 1.2137,
+      "step": 22653
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9734550714492798,
+      "learning_rate": 4.401427438540841e-05,
+      "loss": 1.3239,
+      "step": 22654
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.058668375015259,
+      "learning_rate": 4.401401004493788e-05,
+      "loss": 1.2907,
+      "step": 22655
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2629823684692383,
+      "learning_rate": 4.401374570446736e-05,
+      "loss": 1.3998,
+      "step": 22656
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0433902740478516,
+      "learning_rate": 4.4013481363996826e-05,
+      "loss": 1.3258,
+      "step": 22657
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1357171535491943,
+      "learning_rate": 4.401321702352631e-05,
+      "loss": 1.4756,
+      "step": 22658
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6747477054595947,
+      "learning_rate": 4.401295268305578e-05,
+      "loss": 1.255,
+      "step": 22659
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3880362510681152,
+      "learning_rate": 4.4012688342585254e-05,
+      "loss": 1.3126,
+      "step": 22660
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2890827655792236,
+      "learning_rate": 4.401242400211473e-05,
+      "loss": 1.4267,
+      "step": 22661
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8007007837295532,
+      "learning_rate": 4.40121596616442e-05,
+      "loss": 1.2781,
+      "step": 22662
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8861109018325806,
+      "learning_rate": 4.4011895321173675e-05,
+      "loss": 1.3652,
+      "step": 22663
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8085691928863525,
+      "learning_rate": 4.401163098070315e-05,
+      "loss": 1.1976,
+      "step": 22664
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.5261574983596802,
+      "learning_rate": 4.401136664023262e-05,
+      "loss": 1.1502,
+      "step": 22665
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.642676830291748,
+      "learning_rate": 4.4011102299762095e-05,
+      "loss": 1.2882,
+      "step": 22666
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.9687016010284424,
+      "learning_rate": 4.401083795929157e-05,
+      "loss": 1.3817,
+      "step": 22667
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.478896141052246,
+      "learning_rate": 4.401057361882104e-05,
+      "loss": 1.4761,
+      "step": 22668
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0683352947235107,
+      "learning_rate": 4.401030927835052e-05,
+      "loss": 1.2471,
+      "step": 22669
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.924309015274048,
+      "learning_rate": 4.401004493787999e-05,
+      "loss": 1.4016,
+      "step": 22670
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.034696578979492,
+      "learning_rate": 4.400978059740947e-05,
+      "loss": 1.3944,
+      "step": 22671
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6004639863967896,
+      "learning_rate": 4.400951625693894e-05,
+      "loss": 1.3014,
+      "step": 22672
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7832499742507935,
+      "learning_rate": 4.400925191646841e-05,
+      "loss": 1.1493,
+      "step": 22673
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6060657501220703,
+      "learning_rate": 4.400898757599789e-05,
+      "loss": 1.066,
+      "step": 22674
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1571919918060303,
+      "learning_rate": 4.400872323552736e-05,
+      "loss": 1.0731,
+      "step": 22675
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1696367263793945,
+      "learning_rate": 4.400845889505684e-05,
+      "loss": 1.3691,
+      "step": 22676
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.5792229175567627,
+      "learning_rate": 4.4008194554586305e-05,
+      "loss": 1.2973,
+      "step": 22677
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1520256996154785,
+      "learning_rate": 4.4007930214115785e-05,
+      "loss": 1.2163,
+      "step": 22678
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0027918815612793,
+      "learning_rate": 4.400766587364526e-05,
+      "loss": 1.3561,
+      "step": 22679
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9949519634246826,
+      "learning_rate": 4.400740153317473e-05,
+      "loss": 1.2042,
+      "step": 22680
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9883708953857422,
+      "learning_rate": 4.4007137192704206e-05,
+      "loss": 1.3522,
+      "step": 22681
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.447248935699463,
+      "learning_rate": 4.400687285223368e-05,
+      "loss": 1.4034,
+      "step": 22682
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0173678398132324,
+      "learning_rate": 4.400660851176315e-05,
+      "loss": 1.3222,
+      "step": 22683
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9084967374801636,
+      "learning_rate": 4.4006344171292626e-05,
+      "loss": 1.4156,
+      "step": 22684
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9361521005630493,
+      "learning_rate": 4.40060798308221e-05,
+      "loss": 1.2675,
+      "step": 22685
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2824106216430664,
+      "learning_rate": 4.4005815490351574e-05,
+      "loss": 1.2874,
+      "step": 22686
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4161243438720703,
+      "learning_rate": 4.400555114988105e-05,
+      "loss": 1.2411,
+      "step": 22687
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0440480709075928,
+      "learning_rate": 4.400528680941052e-05,
+      "loss": 1.2406,
+      "step": 22688
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.038451910018921,
+      "learning_rate": 4.400502246894e-05,
+      "loss": 1.3908,
+      "step": 22689
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0573506355285645,
+      "learning_rate": 4.400475812846947e-05,
+      "loss": 1.3322,
+      "step": 22690
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9268518686294556,
+      "learning_rate": 4.400449378799895e-05,
+      "loss": 1.2848,
+      "step": 22691
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.859050989151001,
+      "learning_rate": 4.4004229447528415e-05,
+      "loss": 1.2511,
+      "step": 22692
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.791235089302063,
+      "learning_rate": 4.4003965107057895e-05,
+      "loss": 1.2191,
+      "step": 22693
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7093156576156616,
+      "learning_rate": 4.400370076658737e-05,
+      "loss": 1.2423,
+      "step": 22694
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.601494789123535,
+      "learning_rate": 4.400343642611684e-05,
+      "loss": 1.5049,
+      "step": 22695
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.499426007270813,
+      "learning_rate": 4.4003172085646316e-05,
+      "loss": 1.26,
+      "step": 22696
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8255455493927002,
+      "learning_rate": 4.400290774517579e-05,
+      "loss": 1.9138,
+      "step": 22697
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.981061577796936,
+      "learning_rate": 4.400264340470526e-05,
+      "loss": 2.1498,
+      "step": 22698
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8791794776916504,
+      "learning_rate": 4.400237906423474e-05,
+      "loss": 1.1609,
+      "step": 22699
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.447411060333252,
+      "learning_rate": 4.400211472376421e-05,
+      "loss": 1.2557,
+      "step": 22700
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0240252017974854,
+      "learning_rate": 4.4001850383293684e-05,
+      "loss": 1.4647,
+      "step": 22701
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.902761459350586,
+      "learning_rate": 4.400158604282316e-05,
+      "loss": 1.1906,
+      "step": 22702
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.937820553779602,
+      "learning_rate": 4.400132170235263e-05,
+      "loss": 1.1907,
+      "step": 22703
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.237260341644287,
+      "learning_rate": 4.400105736188211e-05,
+      "loss": 1.4741,
+      "step": 22704
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0397300720214844,
+      "learning_rate": 4.400079302141158e-05,
+      "loss": 1.3411,
+      "step": 22705
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1351382732391357,
+      "learning_rate": 4.400052868094106e-05,
+      "loss": 1.2913,
+      "step": 22706
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8527791500091553,
+      "learning_rate": 4.4000264340470525e-05,
+      "loss": 1.3724,
+      "step": 22707
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1060256958007812,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.417,
+      "step": 22708
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.04010272026062,
+      "learning_rate": 4.399973565952948e-05,
+      "loss": 1.3443,
+      "step": 22709
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.64963698387146,
+      "learning_rate": 4.3999471319058946e-05,
+      "loss": 1.1289,
+      "step": 22710
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8962032794952393,
+      "learning_rate": 4.3999206978588426e-05,
+      "loss": 1.3556,
+      "step": 22711
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8462897539138794,
+      "learning_rate": 4.399894263811789e-05,
+      "loss": 1.4184,
+      "step": 22712
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0189690589904785,
+      "learning_rate": 4.3998678297647373e-05,
+      "loss": 1.2316,
+      "step": 22713
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.689598798751831,
+      "learning_rate": 4.399841395717685e-05,
+      "loss": 1.654,
+      "step": 22714
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8693516254425049,
+      "learning_rate": 4.399814961670632e-05,
+      "loss": 1.334,
+      "step": 22715
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8846193552017212,
+      "learning_rate": 4.3997885276235794e-05,
+      "loss": 1.1025,
+      "step": 22716
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8984675407409668,
+      "learning_rate": 4.399762093576527e-05,
+      "loss": 1.3103,
+      "step": 22717
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.5132020711898804,
+      "learning_rate": 4.399735659529474e-05,
+      "loss": 1.2263,
+      "step": 22718
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8414015769958496,
+      "learning_rate": 4.3997092254824215e-05,
+      "loss": 1.2624,
+      "step": 22719
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.144158363342285,
+      "learning_rate": 4.399682791435369e-05,
+      "loss": 1.3325,
+      "step": 22720
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4051783084869385,
+      "learning_rate": 4.399656357388316e-05,
+      "loss": 1.2723,
+      "step": 22721
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.212185859680176,
+      "learning_rate": 4.3996299233412636e-05,
+      "loss": 1.3546,
+      "step": 22722
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9793871641159058,
+      "learning_rate": 4.399603489294211e-05,
+      "loss": 1.3127,
+      "step": 22723
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0535287857055664,
+      "learning_rate": 4.399577055247159e-05,
+      "loss": 1.1824,
+      "step": 22724
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8253533840179443,
+      "learning_rate": 4.3995506212001056e-05,
+      "loss": 1.2654,
+      "step": 22725
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9827296733856201,
+      "learning_rate": 4.399524187153054e-05,
+      "loss": 1.3117,
+      "step": 22726
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3608036041259766,
+      "learning_rate": 4.3994977531060003e-05,
+      "loss": 1.2941,
+      "step": 22727
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.031299591064453,
+      "learning_rate": 4.3994713190589484e-05,
+      "loss": 1.3608,
+      "step": 22728
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2524795532226562,
+      "learning_rate": 4.399444885011896e-05,
+      "loss": 1.2552,
+      "step": 22729
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9823734760284424,
+      "learning_rate": 4.399418450964843e-05,
+      "loss": 1.2438,
+      "step": 22730
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1481854915618896,
+      "learning_rate": 4.3993920169177905e-05,
+      "loss": 1.347,
+      "step": 22731
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.305650472640991,
+      "learning_rate": 4.399365582870738e-05,
+      "loss": 1.3239,
+      "step": 22732
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8873286247253418,
+      "learning_rate": 4.399339148823685e-05,
+      "loss": 1.4167,
+      "step": 22733
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0200822353363037,
+      "learning_rate": 4.3993127147766325e-05,
+      "loss": 1.3722,
+      "step": 22734
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1837260723114014,
+      "learning_rate": 4.39928628072958e-05,
+      "loss": 1.2919,
+      "step": 22735
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.80424964427948,
+      "learning_rate": 4.399259846682527e-05,
+      "loss": 1.2534,
+      "step": 22736
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6246967315673828,
+      "learning_rate": 4.3992334126354746e-05,
+      "loss": 1.2405,
+      "step": 22737
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0075106620788574,
+      "learning_rate": 4.399206978588422e-05,
+      "loss": 1.2748,
+      "step": 22738
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.5354721546173096,
+      "learning_rate": 4.39918054454137e-05,
+      "loss": 1.2753,
+      "step": 22739
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.248746395111084,
+      "learning_rate": 4.399154110494317e-05,
+      "loss": 1.3762,
+      "step": 22740
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1415674686431885,
+      "learning_rate": 4.399127676447265e-05,
+      "loss": 1.2964,
+      "step": 22741
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1964213848114014,
+      "learning_rate": 4.3991012424002114e-05,
+      "loss": 1.3676,
+      "step": 22742
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.051530599594116,
+      "learning_rate": 4.3990748083531594e-05,
+      "loss": 1.3413,
+      "step": 22743
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.34971022605896,
+      "learning_rate": 4.399048374306107e-05,
+      "loss": 1.3574,
+      "step": 22744
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8878992795944214,
+      "learning_rate": 4.3990219402590535e-05,
+      "loss": 1.2716,
+      "step": 22745
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8259683847427368,
+      "learning_rate": 4.3989955062120015e-05,
+      "loss": 1.3324,
+      "step": 22746
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.792876958847046,
+      "learning_rate": 4.398969072164948e-05,
+      "loss": 1.2496,
+      "step": 22747
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.06848406791687,
+      "learning_rate": 4.398942638117896e-05,
+      "loss": 1.422,
+      "step": 22748
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.581636667251587,
+      "learning_rate": 4.3989162040708436e-05,
+      "loss": 1.4401,
+      "step": 22749
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6939789056777954,
+      "learning_rate": 4.398889770023791e-05,
+      "loss": 1.2477,
+      "step": 22750
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9005770683288574,
+      "learning_rate": 4.398863335976738e-05,
+      "loss": 1.219,
+      "step": 22751
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.5139689445495605,
+      "learning_rate": 4.3988369019296856e-05,
+      "loss": 1.2291,
+      "step": 22752
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2135136127471924,
+      "learning_rate": 4.398810467882633e-05,
+      "loss": 1.3466,
+      "step": 22753
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0155105590820312,
+      "learning_rate": 4.398784033835581e-05,
+      "loss": 1.3943,
+      "step": 22754
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.056762456893921,
+      "learning_rate": 4.398757599788528e-05,
+      "loss": 1.2937,
+      "step": 22755
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4626994132995605,
+      "learning_rate": 4.398731165741475e-05,
+      "loss": 1.989,
+      "step": 22756
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8491452932357788,
+      "learning_rate": 4.3987047316944224e-05,
+      "loss": 1.2512,
+      "step": 22757
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.076364040374756,
+      "learning_rate": 4.39867829764737e-05,
+      "loss": 1.3646,
+      "step": 22758
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6670753955841064,
+      "learning_rate": 4.398651863600318e-05,
+      "loss": 1.0573,
+      "step": 22759
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4884932041168213,
+      "learning_rate": 4.3986254295532645e-05,
+      "loss": 1.3672,
+      "step": 22760
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.531756401062012,
+      "learning_rate": 4.3985989955062125e-05,
+      "loss": 1.5058,
+      "step": 22761
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.738776445388794,
+      "learning_rate": 4.398572561459159e-05,
+      "loss": 1.301,
+      "step": 22762
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.106502056121826,
+      "learning_rate": 4.398546127412107e-05,
+      "loss": 1.4038,
+      "step": 22763
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8466176986694336,
+      "learning_rate": 4.3985196933650546e-05,
+      "loss": 1.0759,
+      "step": 22764
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7493109703063965,
+      "learning_rate": 4.398493259318002e-05,
+      "loss": 1.2184,
+      "step": 22765
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3172967433929443,
+      "learning_rate": 4.398466825270949e-05,
+      "loss": 1.3714,
+      "step": 22766
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9625450372695923,
+      "learning_rate": 4.398440391223897e-05,
+      "loss": 1.3798,
+      "step": 22767
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2606117725372314,
+      "learning_rate": 4.398413957176844e-05,
+      "loss": 1.3817,
+      "step": 22768
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.071431875228882,
+      "learning_rate": 4.3983875231297914e-05,
+      "loss": 1.4413,
+      "step": 22769
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7811894416809082,
+      "learning_rate": 4.398361089082739e-05,
+      "loss": 1.3529,
+      "step": 22770
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9410828351974487,
+      "learning_rate": 4.398334655035686e-05,
+      "loss": 1.3254,
+      "step": 22771
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.893492341041565,
+      "learning_rate": 4.3983082209886334e-05,
+      "loss": 1.3786,
+      "step": 22772
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.5252366065979004,
+      "learning_rate": 4.398281786941581e-05,
+      "loss": 1.4786,
+      "step": 22773
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.9029183387756348,
+      "learning_rate": 4.398255352894529e-05,
+      "loss": 1.2242,
+      "step": 22774
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8094481229782104,
+      "learning_rate": 4.3982289188474755e-05,
+      "loss": 1.2163,
+      "step": 22775
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.6850814819335938,
+      "learning_rate": 4.3982024848004236e-05,
+      "loss": 1.2334,
+      "step": 22776
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3517005443573,
+      "learning_rate": 4.39817605075337e-05,
+      "loss": 1.2892,
+      "step": 22777
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.6693968772888184,
+      "learning_rate": 4.398149616706318e-05,
+      "loss": 2.032,
+      "step": 22778
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.672648310661316,
+      "learning_rate": 4.3981231826592656e-05,
+      "loss": 1.1605,
+      "step": 22779
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.389989137649536,
+      "learning_rate": 4.398096748612213e-05,
+      "loss": 1.3672,
+      "step": 22780
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.333437442779541,
+      "learning_rate": 4.3980703145651603e-05,
+      "loss": 1.3877,
+      "step": 22781
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.325638771057129,
+      "learning_rate": 4.398043880518107e-05,
+      "loss": 1.4318,
+      "step": 22782
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0692858695983887,
+      "learning_rate": 4.398017446471055e-05,
+      "loss": 1.3255,
+      "step": 22783
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.6160671710968018,
+      "learning_rate": 4.3979910124240024e-05,
+      "loss": 1.3817,
+      "step": 22784
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.343574047088623,
+      "learning_rate": 4.39796457837695e-05,
+      "loss": 1.3791,
+      "step": 22785
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.442366123199463,
+      "learning_rate": 4.397938144329897e-05,
+      "loss": 1.3393,
+      "step": 22786
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.322824001312256,
+      "learning_rate": 4.3979117102828445e-05,
+      "loss": 1.2663,
+      "step": 22787
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7971084117889404,
+      "learning_rate": 4.397885276235792e-05,
+      "loss": 1.2747,
+      "step": 22788
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.102074146270752,
+      "learning_rate": 4.39785884218874e-05,
+      "loss": 1.3932,
+      "step": 22789
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.9720592498779297,
+      "learning_rate": 4.3978324081416866e-05,
+      "loss": 2.1642,
+      "step": 22790
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0161972045898438,
+      "learning_rate": 4.397805974094634e-05,
+      "loss": 1.222,
+      "step": 22791
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.002793788909912,
+      "learning_rate": 4.397779540047581e-05,
+      "loss": 1.2094,
+      "step": 22792
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.198802947998047,
+      "learning_rate": 4.3977531060005286e-05,
+      "loss": 1.4487,
+      "step": 22793
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.1039535999298096,
+      "learning_rate": 4.397726671953477e-05,
+      "loss": 1.4029,
+      "step": 22794
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.704229474067688,
+      "learning_rate": 4.3977002379064233e-05,
+      "loss": 1.18,
+      "step": 22795
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8552381992340088,
+      "learning_rate": 4.3976738038593714e-05,
+      "loss": 1.3894,
+      "step": 22796
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0953145027160645,
+      "learning_rate": 4.397647369812318e-05,
+      "loss": 1.4028,
+      "step": 22797
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.173906087875366,
+      "learning_rate": 4.397620935765266e-05,
+      "loss": 1.254,
+      "step": 22798
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1348824501037598,
+      "learning_rate": 4.3975945017182134e-05,
+      "loss": 1.3625,
+      "step": 22799
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.171196460723877,
+      "learning_rate": 4.397568067671161e-05,
+      "loss": 1.2558,
+      "step": 22800
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9188504219055176,
+      "learning_rate": 4.397541633624108e-05,
+      "loss": 1.1863,
+      "step": 22801
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7145202159881592,
+      "learning_rate": 4.3975151995770555e-05,
+      "loss": 1.2048,
+      "step": 22802
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.264878273010254,
+      "learning_rate": 4.397488765530003e-05,
+      "loss": 1.245,
+      "step": 22803
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.828275203704834,
+      "learning_rate": 4.39746233148295e-05,
+      "loss": 1.477,
+      "step": 22804
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8370956182479858,
+      "learning_rate": 4.3974358974358976e-05,
+      "loss": 1.6169,
+      "step": 22805
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.052962303161621,
+      "learning_rate": 4.397409463388845e-05,
+      "loss": 1.3047,
+      "step": 22806
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.186457633972168,
+      "learning_rate": 4.397383029341792e-05,
+      "loss": 1.2365,
+      "step": 22807
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.111788034439087,
+      "learning_rate": 4.3973565952947397e-05,
+      "loss": 1.4164,
+      "step": 22808
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0005829334259033,
+      "learning_rate": 4.397330161247688e-05,
+      "loss": 1.3201,
+      "step": 22809
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.026567220687866,
+      "learning_rate": 4.3973037272006344e-05,
+      "loss": 1.3452,
+      "step": 22810
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.035167932510376,
+      "learning_rate": 4.3972772931535824e-05,
+      "loss": 1.4135,
+      "step": 22811
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9576106071472168,
+      "learning_rate": 4.397250859106529e-05,
+      "loss": 1.1738,
+      "step": 22812
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9624512195587158,
+      "learning_rate": 4.397224425059477e-05,
+      "loss": 1.3527,
+      "step": 22813
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.0738296508789062,
+      "learning_rate": 4.3971979910124245e-05,
+      "loss": 1.2272,
+      "step": 22814
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.952452540397644,
+      "learning_rate": 4.397171556965372e-05,
+      "loss": 1.363,
+      "step": 22815
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1937062740325928,
+      "learning_rate": 4.397145122918319e-05,
+      "loss": 1.1951,
+      "step": 22816
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7110536098480225,
+      "learning_rate": 4.397118688871266e-05,
+      "loss": 1.2677,
+      "step": 22817
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.158071517944336,
+      "learning_rate": 4.397092254824214e-05,
+      "loss": 1.1607,
+      "step": 22818
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.931880235671997,
+      "learning_rate": 4.397065820777161e-05,
+      "loss": 1.2809,
+      "step": 22819
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.3105859756469727,
+      "learning_rate": 4.3970393867301086e-05,
+      "loss": 1.4606,
+      "step": 22820
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.4557883739471436,
+      "learning_rate": 4.397012952683056e-05,
+      "loss": 1.4381,
+      "step": 22821
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8494465351104736,
+      "learning_rate": 4.396986518636003e-05,
+      "loss": 1.4124,
+      "step": 22822
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.798696756362915,
+      "learning_rate": 4.396960084588951e-05,
+      "loss": 1.3953,
+      "step": 22823
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.268935441970825,
+      "learning_rate": 4.396933650541899e-05,
+      "loss": 1.3261,
+      "step": 22824
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.946386694908142,
+      "learning_rate": 4.3969072164948454e-05,
+      "loss": 1.4243,
+      "step": 22825
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0125770568847656,
+      "learning_rate": 4.3968807824477934e-05,
+      "loss": 1.2499,
+      "step": 22826
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0286173820495605,
+      "learning_rate": 4.39685434840074e-05,
+      "loss": 1.2983,
+      "step": 22827
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.767219066619873,
+      "learning_rate": 4.3968279143536875e-05,
+      "loss": 1.2577,
+      "step": 22828
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9991683959960938,
+      "learning_rate": 4.3968014803066355e-05,
+      "loss": 1.3018,
+      "step": 22829
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.104456901550293,
+      "learning_rate": 4.396775046259582e-05,
+      "loss": 1.3013,
+      "step": 22830
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0365002155303955,
+      "learning_rate": 4.39674861221253e-05,
+      "loss": 1.324,
+      "step": 22831
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.5845847129821777,
+      "learning_rate": 4.396722178165477e-05,
+      "loss": 1.2989,
+      "step": 22832
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.47598934173584,
+      "learning_rate": 4.396695744118425e-05,
+      "loss": 1.4763,
+      "step": 22833
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0884501934051514,
+      "learning_rate": 4.396669310071372e-05,
+      "loss": 1.2349,
+      "step": 22834
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8319216966629028,
+      "learning_rate": 4.3966428760243197e-05,
+      "loss": 1.4163,
+      "step": 22835
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9600855112075806,
+      "learning_rate": 4.396616441977267e-05,
+      "loss": 1.4334,
+      "step": 22836
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2542850971221924,
+      "learning_rate": 4.3965900079302144e-05,
+      "loss": 1.2067,
+      "step": 22837
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8019968271255493,
+      "learning_rate": 4.396563573883162e-05,
+      "loss": 1.3165,
+      "step": 22838
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1772301197052,
+      "learning_rate": 4.396537139836109e-05,
+      "loss": 1.1392,
+      "step": 22839
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.266885995864868,
+      "learning_rate": 4.3965107057890564e-05,
+      "loss": 1.3146,
+      "step": 22840
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.734601378440857,
+      "learning_rate": 4.396484271742004e-05,
+      "loss": 1.236,
+      "step": 22841
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0513200759887695,
+      "learning_rate": 4.396457837694951e-05,
+      "loss": 1.252,
+      "step": 22842
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2370431423187256,
+      "learning_rate": 4.3964314036478985e-05,
+      "loss": 1.4977,
+      "step": 22843
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9637067317962646,
+      "learning_rate": 4.3964049696008465e-05,
+      "loss": 1.36,
+      "step": 22844
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.812084436416626,
+      "learning_rate": 4.396378535553793e-05,
+      "loss": 1.3536,
+      "step": 22845
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.919795274734497,
+      "learning_rate": 4.396352101506741e-05,
+      "loss": 1.2027,
+      "step": 22846
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.49946665763855,
+      "learning_rate": 4.396325667459688e-05,
+      "loss": 1.5457,
+      "step": 22847
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.030522584915161,
+      "learning_rate": 4.396299233412636e-05,
+      "loss": 1.428,
+      "step": 22848
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7661818265914917,
+      "learning_rate": 4.396272799365583e-05,
+      "loss": 1.2438,
+      "step": 22849
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9421560764312744,
+      "learning_rate": 4.396246365318531e-05,
+      "loss": 1.2085,
+      "step": 22850
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.5255396366119385,
+      "learning_rate": 4.396219931271478e-05,
+      "loss": 1.2909,
+      "step": 22851
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9818806648254395,
+      "learning_rate": 4.3961934972244254e-05,
+      "loss": 1.2989,
+      "step": 22852
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.597870945930481,
+      "learning_rate": 4.396167063177373e-05,
+      "loss": 1.2441,
+      "step": 22853
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0709099769592285,
+      "learning_rate": 4.39614062913032e-05,
+      "loss": 1.3825,
+      "step": 22854
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1012840270996094,
+      "learning_rate": 4.3961141950832675e-05,
+      "loss": 1.4106,
+      "step": 22855
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0840859413146973,
+      "learning_rate": 4.396087761036215e-05,
+      "loss": 1.3435,
+      "step": 22856
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.02233624458313,
+      "learning_rate": 4.396061326989162e-05,
+      "loss": 1.3,
+      "step": 22857
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.686391830444336,
+      "learning_rate": 4.3960348929421095e-05,
+      "loss": 1.2567,
+      "step": 22858
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2629013061523438,
+      "learning_rate": 4.3960084588950576e-05,
+      "loss": 1.2975,
+      "step": 22859
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1783556938171387,
+      "learning_rate": 4.395982024848004e-05,
+      "loss": 1.3113,
+      "step": 22860
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1946675777435303,
+      "learning_rate": 4.395955590800952e-05,
+      "loss": 1.4038,
+      "step": 22861
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.820923924446106,
+      "learning_rate": 4.395929156753899e-05,
+      "loss": 1.3623,
+      "step": 22862
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9010599851608276,
+      "learning_rate": 4.395902722706846e-05,
+      "loss": 1.4156,
+      "step": 22863
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.6177778244018555,
+      "learning_rate": 4.3958762886597944e-05,
+      "loss": 1.4692,
+      "step": 22864
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.6401941776275635,
+      "learning_rate": 4.395849854612741e-05,
+      "loss": 1.3039,
+      "step": 22865
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.661313056945801,
+      "learning_rate": 4.395823420565689e-05,
+      "loss": 1.3473,
+      "step": 22866
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7651901245117188,
+      "learning_rate": 4.395796986518636e-05,
+      "loss": 1.2063,
+      "step": 22867
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.079261064529419,
+      "learning_rate": 4.395770552471584e-05,
+      "loss": 1.276,
+      "step": 22868
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.78025221824646,
+      "learning_rate": 4.395744118424531e-05,
+      "loss": 1.1831,
+      "step": 22869
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.7319188117980957,
+      "learning_rate": 4.3957176843774785e-05,
+      "loss": 1.4747,
+      "step": 22870
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1259102821350098,
+      "learning_rate": 4.395691250330426e-05,
+      "loss": 1.2659,
+      "step": 22871
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.869295597076416,
+      "learning_rate": 4.395664816283373e-05,
+      "loss": 1.3145,
+      "step": 22872
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.4300384521484375,
+      "learning_rate": 4.3956383822363206e-05,
+      "loss": 3.9217,
+      "step": 22873
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.99752938747406,
+      "learning_rate": 4.395611948189268e-05,
+      "loss": 1.4879,
+      "step": 22874
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.402158260345459,
+      "learning_rate": 4.395585514142215e-05,
+      "loss": 1.3337,
+      "step": 22875
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.6684619188308716,
+      "learning_rate": 4.3955590800951627e-05,
+      "loss": 1.5672,
+      "step": 22876
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2801125049591064,
+      "learning_rate": 4.39553264604811e-05,
+      "loss": 1.3481,
+      "step": 22877
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.070566415786743,
+      "learning_rate": 4.3955062120010574e-05,
+      "loss": 1.4346,
+      "step": 22878
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1374447345733643,
+      "learning_rate": 4.3954797779540054e-05,
+      "loss": 1.3233,
+      "step": 22879
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.09423565864563,
+      "learning_rate": 4.395453343906952e-05,
+      "loss": 1.2957,
+      "step": 22880
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8014941215515137,
+      "learning_rate": 4.3954269098599e-05,
+      "loss": 1.184,
+      "step": 22881
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9338419437408447,
+      "learning_rate": 4.395400475812847e-05,
+      "loss": 1.3378,
+      "step": 22882
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.5385758876800537,
+      "learning_rate": 4.395374041765795e-05,
+      "loss": 1.1513,
+      "step": 22883
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.723767638206482,
+      "learning_rate": 4.395347607718742e-05,
+      "loss": 1.1946,
+      "step": 22884
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8939167261123657,
+      "learning_rate": 4.3953211736716895e-05,
+      "loss": 1.2731,
+      "step": 22885
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.045602798461914,
+      "learning_rate": 4.395294739624637e-05,
+      "loss": 1.2432,
+      "step": 22886
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.254286766052246,
+      "learning_rate": 4.395268305577584e-05,
+      "loss": 1.4961,
+      "step": 22887
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.179844856262207,
+      "learning_rate": 4.3952418715305316e-05,
+      "loss": 1.3055,
+      "step": 22888
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8152116537094116,
+      "learning_rate": 4.395215437483479e-05,
+      "loss": 1.3101,
+      "step": 22889
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0464119911193848,
+      "learning_rate": 4.395189003436426e-05,
+      "loss": 1.3182,
+      "step": 22890
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8477805852890015,
+      "learning_rate": 4.395162569389374e-05,
+      "loss": 1.4075,
+      "step": 22891
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.556490421295166,
+      "learning_rate": 4.395136135342321e-05,
+      "loss": 1.32,
+      "step": 22892
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.852448582649231,
+      "learning_rate": 4.3951097012952684e-05,
+      "loss": 1.1738,
+      "step": 22893
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.1450769901275635,
+      "learning_rate": 4.3950832672482164e-05,
+      "loss": 1.4954,
+      "step": 22894
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.4139959812164307,
+      "learning_rate": 4.395056833201163e-05,
+      "loss": 1.3875,
+      "step": 22895
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.0349225997924805,
+      "learning_rate": 4.395030399154111e-05,
+      "loss": 1.2416,
+      "step": 22896
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.6755889654159546,
+      "learning_rate": 4.395003965107058e-05,
+      "loss": 1.1213,
+      "step": 22897
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.008084535598755,
+      "learning_rate": 4.394977531060006e-05,
+      "loss": 1.235,
+      "step": 22898
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9526017904281616,
+      "learning_rate": 4.394951097012953e-05,
+      "loss": 1.2561,
+      "step": 22899
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1671862602233887,
+      "learning_rate": 4.3949246629659e-05,
+      "loss": 1.374,
+      "step": 22900
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9623408317565918,
+      "learning_rate": 4.394898228918848e-05,
+      "loss": 1.2666,
+      "step": 22901
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.6391563415527344,
+      "learning_rate": 4.3948717948717946e-05,
+      "loss": 1.4993,
+      "step": 22902
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.396418333053589,
+      "learning_rate": 4.3948453608247427e-05,
+      "loss": 1.2642,
+      "step": 22903
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8872445821762085,
+      "learning_rate": 4.39481892677769e-05,
+      "loss": 1.2489,
+      "step": 22904
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.257706880569458,
+      "learning_rate": 4.3947924927306374e-05,
+      "loss": 1.2777,
+      "step": 22905
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.599181890487671,
+      "learning_rate": 4.394766058683585e-05,
+      "loss": 1.1463,
+      "step": 22906
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.4454352855682373,
+      "learning_rate": 4.394739624636532e-05,
+      "loss": 1.286,
+      "step": 22907
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.03059720993042,
+      "learning_rate": 4.3947131905894794e-05,
+      "loss": 1.3454,
+      "step": 22908
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9756450653076172,
+      "learning_rate": 4.394686756542427e-05,
+      "loss": 1.3219,
+      "step": 22909
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2088944911956787,
+      "learning_rate": 4.394660322495374e-05,
+      "loss": 1.378,
+      "step": 22910
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9868144989013672,
+      "learning_rate": 4.3946338884483215e-05,
+      "loss": 1.4057,
+      "step": 22911
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8301202058792114,
+      "learning_rate": 4.394607454401269e-05,
+      "loss": 1.2074,
+      "step": 22912
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8907191753387451,
+      "learning_rate": 4.394581020354216e-05,
+      "loss": 1.6557,
+      "step": 22913
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.5533627271652222,
+      "learning_rate": 4.394554586307164e-05,
+      "loss": 1.1646,
+      "step": 22914
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2478132247924805,
+      "learning_rate": 4.394528152260111e-05,
+      "loss": 1.4158,
+      "step": 22915
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0211236476898193,
+      "learning_rate": 4.394501718213059e-05,
+      "loss": 1.8383,
+      "step": 22916
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.486112356185913,
+      "learning_rate": 4.3944752841660056e-05,
+      "loss": 1.3151,
+      "step": 22917
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.5990114212036133,
+      "learning_rate": 4.394448850118954e-05,
+      "loss": 1.3809,
+      "step": 22918
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9728199243545532,
+      "learning_rate": 4.394422416071901e-05,
+      "loss": 1.3534,
+      "step": 22919
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7582638263702393,
+      "learning_rate": 4.3943959820248484e-05,
+      "loss": 1.1699,
+      "step": 22920
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.699397087097168,
+      "learning_rate": 4.394369547977796e-05,
+      "loss": 1.4532,
+      "step": 22921
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.536040782928467,
+      "learning_rate": 4.394343113930743e-05,
+      "loss": 1.1341,
+      "step": 22922
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.170426845550537,
+      "learning_rate": 4.3943166798836905e-05,
+      "loss": 1.2541,
+      "step": 22923
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7444897890090942,
+      "learning_rate": 4.394290245836638e-05,
+      "loss": 1.9276,
+      "step": 22924
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0301384925842285,
+      "learning_rate": 4.394263811789585e-05,
+      "loss": 1.4074,
+      "step": 22925
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.649073839187622,
+      "learning_rate": 4.3942373777425325e-05,
+      "loss": 1.3771,
+      "step": 22926
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.6930956840515137,
+      "learning_rate": 4.39421094369548e-05,
+      "loss": 1.4105,
+      "step": 22927
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.938735008239746,
+      "learning_rate": 4.394184509648427e-05,
+      "loss": 1.8594,
+      "step": 22928
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.094564199447632,
+      "learning_rate": 4.394158075601375e-05,
+      "loss": 1.3696,
+      "step": 22929
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.6626052856445312,
+      "learning_rate": 4.394131641554322e-05,
+      "loss": 1.3738,
+      "step": 22930
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.180739164352417,
+      "learning_rate": 4.39410520750727e-05,
+      "loss": 1.3915,
+      "step": 22931
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.4058992862701416,
+      "learning_rate": 4.394078773460217e-05,
+      "loss": 1.362,
+      "step": 22932
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.6655335426330566,
+      "learning_rate": 4.394052339413165e-05,
+      "loss": 2.0789,
+      "step": 22933
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0842349529266357,
+      "learning_rate": 4.394025905366112e-05,
+      "loss": 1.3322,
+      "step": 22934
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.848280191421509,
+      "learning_rate": 4.393999471319059e-05,
+      "loss": 1.4031,
+      "step": 22935
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1078732013702393,
+      "learning_rate": 4.393973037272007e-05,
+      "loss": 1.2178,
+      "step": 22936
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.5952348709106445,
+      "learning_rate": 4.3939466032249535e-05,
+      "loss": 1.4066,
+      "step": 22937
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.5883593559265137,
+      "learning_rate": 4.3939201691779015e-05,
+      "loss": 1.3567,
+      "step": 22938
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0708417892456055,
+      "learning_rate": 4.393893735130849e-05,
+      "loss": 1.275,
+      "step": 22939
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.755903959274292,
+      "learning_rate": 4.393867301083796e-05,
+      "loss": 1.2844,
+      "step": 22940
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9586116075515747,
+      "learning_rate": 4.3938408670367436e-05,
+      "loss": 1.0889,
+      "step": 22941
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1447277069091797,
+      "learning_rate": 4.393814432989691e-05,
+      "loss": 1.4184,
+      "step": 22942
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.5390591621398926,
+      "learning_rate": 4.393787998942638e-05,
+      "loss": 1.3052,
+      "step": 22943
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.912253737449646,
+      "learning_rate": 4.393761564895586e-05,
+      "loss": 1.2688,
+      "step": 22944
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9469804763793945,
+      "learning_rate": 4.393735130848533e-05,
+      "loss": 1.2783,
+      "step": 22945
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.848840594291687,
+      "learning_rate": 4.3937086968014804e-05,
+      "loss": 1.3011,
+      "step": 22946
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.845432996749878,
+      "learning_rate": 4.393682262754428e-05,
+      "loss": 1.2964,
+      "step": 22947
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.542739152908325,
+      "learning_rate": 4.393655828707375e-05,
+      "loss": 1.4464,
+      "step": 22948
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0900511741638184,
+      "learning_rate": 4.393629394660323e-05,
+      "loss": 1.3318,
+      "step": 22949
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.91439688205719,
+      "learning_rate": 4.39360296061327e-05,
+      "loss": 1.2094,
+      "step": 22950
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8209810256958008,
+      "learning_rate": 4.393576526566218e-05,
+      "loss": 1.1928,
+      "step": 22951
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1446661949157715,
+      "learning_rate": 4.3935500925191645e-05,
+      "loss": 1.0963,
+      "step": 22952
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0607879161834717,
+      "learning_rate": 4.3935236584721125e-05,
+      "loss": 1.3274,
+      "step": 22953
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7097055912017822,
+      "learning_rate": 4.39349722442506e-05,
+      "loss": 1.3225,
+      "step": 22954
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0612783432006836,
+      "learning_rate": 4.393470790378007e-05,
+      "loss": 1.4439,
+      "step": 22955
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.552689790725708,
+      "learning_rate": 4.3934443563309546e-05,
+      "loss": 1.3205,
+      "step": 22956
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.067078113555908,
+      "learning_rate": 4.393417922283902e-05,
+      "loss": 1.1886,
+      "step": 22957
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.925386667251587,
+      "learning_rate": 4.393391488236849e-05,
+      "loss": 1.485,
+      "step": 22958
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.357356548309326,
+      "learning_rate": 4.393365054189797e-05,
+      "loss": 1.5017,
+      "step": 22959
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0931034088134766,
+      "learning_rate": 4.393338620142744e-05,
+      "loss": 1.3072,
+      "step": 22960
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0553104877471924,
+      "learning_rate": 4.3933121860956914e-05,
+      "loss": 1.4587,
+      "step": 22961
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2752630710601807,
+      "learning_rate": 4.393285752048639e-05,
+      "loss": 1.3848,
+      "step": 22962
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.938120722770691,
+      "learning_rate": 4.393259318001586e-05,
+      "loss": 1.4745,
+      "step": 22963
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.295525074005127,
+      "learning_rate": 4.393232883954534e-05,
+      "loss": 1.3954,
+      "step": 22964
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8941965103149414,
+      "learning_rate": 4.393206449907481e-05,
+      "loss": 1.6625,
+      "step": 22965
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2199325561523438,
+      "learning_rate": 4.393180015860429e-05,
+      "loss": 1.4254,
+      "step": 22966
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.749678611755371,
+      "learning_rate": 4.3931535818133755e-05,
+      "loss": 1.2754,
+      "step": 22967
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2092151641845703,
+      "learning_rate": 4.3931271477663236e-05,
+      "loss": 1.2914,
+      "step": 22968
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9972537755966187,
+      "learning_rate": 4.393100713719271e-05,
+      "loss": 1.2311,
+      "step": 22969
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0343775749206543,
+      "learning_rate": 4.393074279672218e-05,
+      "loss": 1.2069,
+      "step": 22970
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.950446605682373,
+      "learning_rate": 4.3930478456251656e-05,
+      "loss": 1.1997,
+      "step": 22971
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1246464252471924,
+      "learning_rate": 4.393021411578112e-05,
+      "loss": 1.378,
+      "step": 22972
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8270074129104614,
+      "learning_rate": 4.3929949775310604e-05,
+      "loss": 1.3096,
+      "step": 22973
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1470093727111816,
+      "learning_rate": 4.392968543484008e-05,
+      "loss": 1.1179,
+      "step": 22974
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8678103685379028,
+      "learning_rate": 4.392942109436955e-05,
+      "loss": 1.2911,
+      "step": 22975
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1255671977996826,
+      "learning_rate": 4.3929156753899024e-05,
+      "loss": 1.4996,
+      "step": 22976
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0547268390655518,
+      "learning_rate": 4.39288924134285e-05,
+      "loss": 1.1232,
+      "step": 22977
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.0472044944763184,
+      "learning_rate": 4.392862807295797e-05,
+      "loss": 1.2917,
+      "step": 22978
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.701423168182373,
+      "learning_rate": 4.392836373248745e-05,
+      "loss": 1.1674,
+      "step": 22979
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.050194501876831,
+      "learning_rate": 4.392809939201692e-05,
+      "loss": 1.3125,
+      "step": 22980
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.018235445022583,
+      "learning_rate": 4.392783505154639e-05,
+      "loss": 1.2458,
+      "step": 22981
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.3467867374420166,
+      "learning_rate": 4.3927570711075866e-05,
+      "loss": 1.4155,
+      "step": 22982
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.048184871673584,
+      "learning_rate": 4.392730637060534e-05,
+      "loss": 1.2831,
+      "step": 22983
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.29087233543396,
+      "learning_rate": 4.392704203013482e-05,
+      "loss": 1.3992,
+      "step": 22984
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.330232858657837,
+      "learning_rate": 4.3926777689664286e-05,
+      "loss": 1.3332,
+      "step": 22985
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2146151065826416,
+      "learning_rate": 4.392651334919377e-05,
+      "loss": 1.3754,
+      "step": 22986
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.693995714187622,
+      "learning_rate": 4.3926249008723234e-05,
+      "loss": 1.1408,
+      "step": 22987
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7555698156356812,
+      "learning_rate": 4.3925984668252714e-05,
+      "loss": 1.3313,
+      "step": 22988
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.7780191898345947,
+      "learning_rate": 4.392572032778219e-05,
+      "loss": 1.3599,
+      "step": 22989
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2932209968566895,
+      "learning_rate": 4.392545598731166e-05,
+      "loss": 1.2532,
+      "step": 22990
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.4714255332946777,
+      "learning_rate": 4.3925191646841135e-05,
+      "loss": 1.4777,
+      "step": 22991
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.156634569168091,
+      "learning_rate": 4.392492730637061e-05,
+      "loss": 1.3689,
+      "step": 22992
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.090263605117798,
+      "learning_rate": 4.392466296590008e-05,
+      "loss": 1.3898,
+      "step": 22993
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.762970447540283,
+      "learning_rate": 4.3924398625429555e-05,
+      "loss": 1.5744,
+      "step": 22994
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6883050203323364,
+      "learning_rate": 4.392413428495903e-05,
+      "loss": 1.2491,
+      "step": 22995
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.174600601196289,
+      "learning_rate": 4.39238699444885e-05,
+      "loss": 1.4294,
+      "step": 22996
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.8323564529418945,
+      "learning_rate": 4.3923605604017976e-05,
+      "loss": 1.3935,
+      "step": 22997
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.80209481716156,
+      "learning_rate": 4.392334126354745e-05,
+      "loss": 1.3387,
+      "step": 22998
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.29034161567688,
+      "learning_rate": 4.392307692307693e-05,
+      "loss": 1.3372,
+      "step": 22999
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7663780450820923,
+      "learning_rate": 4.39228125826064e-05,
+      "loss": 1.223,
+      "step": 23000
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.969467282295227,
+      "learning_rate": 4.392254824213588e-05,
+      "loss": 1.122,
+      "step": 23001
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.4399025440216064,
+      "learning_rate": 4.3922283901665344e-05,
+      "loss": 1.5613,
+      "step": 23002
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.629751682281494,
+      "learning_rate": 4.3922019561194824e-05,
+      "loss": 1.2125,
+      "step": 23003
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.088818073272705,
+      "learning_rate": 4.39217552207243e-05,
+      "loss": 1.3316,
+      "step": 23004
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2911503314971924,
+      "learning_rate": 4.392149088025377e-05,
+      "loss": 1.3621,
+      "step": 23005
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.821380376815796,
+      "learning_rate": 4.3921226539783245e-05,
+      "loss": 1.2294,
+      "step": 23006
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.841615915298462,
+      "learning_rate": 4.392096219931272e-05,
+      "loss": 1.3533,
+      "step": 23007
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1324706077575684,
+      "learning_rate": 4.392069785884219e-05,
+      "loss": 1.3664,
+      "step": 23008
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6993037462234497,
+      "learning_rate": 4.3920433518371666e-05,
+      "loss": 1.3216,
+      "step": 23009
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9447559118270874,
+      "learning_rate": 4.392016917790114e-05,
+      "loss": 1.3938,
+      "step": 23010
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0244500637054443,
+      "learning_rate": 4.391990483743061e-05,
+      "loss": 1.2928,
+      "step": 23011
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7696304321289062,
+      "learning_rate": 4.3919640496960086e-05,
+      "loss": 1.2943,
+      "step": 23012
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.504415273666382,
+      "learning_rate": 4.391937615648956e-05,
+      "loss": 1.2689,
+      "step": 23013
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7780463695526123,
+      "learning_rate": 4.391911181601904e-05,
+      "loss": 1.325,
+      "step": 23014
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.103895902633667,
+      "learning_rate": 4.391884747554851e-05,
+      "loss": 1.1405,
+      "step": 23015
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6453142166137695,
+      "learning_rate": 4.391858313507799e-05,
+      "loss": 1.182,
+      "step": 23016
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.730010986328125,
+      "learning_rate": 4.3918318794607454e-05,
+      "loss": 1.2978,
+      "step": 23017
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8321894407272339,
+      "learning_rate": 4.391805445413693e-05,
+      "loss": 1.3348,
+      "step": 23018
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7148326635360718,
+      "learning_rate": 4.391779011366641e-05,
+      "loss": 1.7959,
+      "step": 23019
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.424666166305542,
+      "learning_rate": 4.3917525773195875e-05,
+      "loss": 1.3955,
+      "step": 23020
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.408111095428467,
+      "learning_rate": 4.3917261432725355e-05,
+      "loss": 1.3886,
+      "step": 23021
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9382562637329102,
+      "learning_rate": 4.391699709225482e-05,
+      "loss": 1.3951,
+      "step": 23022
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.459824562072754,
+      "learning_rate": 4.39167327517843e-05,
+      "loss": 1.3391,
+      "step": 23023
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.5611488819122314,
+      "learning_rate": 4.3916468411313776e-05,
+      "loss": 1.1486,
+      "step": 23024
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7684171199798584,
+      "learning_rate": 4.391620407084325e-05,
+      "loss": 1.2332,
+      "step": 23025
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3835999965667725,
+      "learning_rate": 4.391593973037272e-05,
+      "loss": 1.3195,
+      "step": 23026
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9038645029067993,
+      "learning_rate": 4.39156753899022e-05,
+      "loss": 1.2776,
+      "step": 23027
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0842812061309814,
+      "learning_rate": 4.391541104943167e-05,
+      "loss": 1.1669,
+      "step": 23028
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.704148769378662,
+      "learning_rate": 4.3915146708961144e-05,
+      "loss": 1.1245,
+      "step": 23029
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7836201190948486,
+      "learning_rate": 4.391488236849062e-05,
+      "loss": 1.27,
+      "step": 23030
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.011802911758423,
+      "learning_rate": 4.391461802802009e-05,
+      "loss": 1.2354,
+      "step": 23031
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.847917079925537,
+      "learning_rate": 4.3914353687549565e-05,
+      "loss": 1.2705,
+      "step": 23032
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.361956834793091,
+      "learning_rate": 4.391408934707904e-05,
+      "loss": 1.4045,
+      "step": 23033
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2575056552886963,
+      "learning_rate": 4.391382500660852e-05,
+      "loss": 1.4335,
+      "step": 23034
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9016183614730835,
+      "learning_rate": 4.3913560666137985e-05,
+      "loss": 1.2735,
+      "step": 23035
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9448821544647217,
+      "learning_rate": 4.3913296325667466e-05,
+      "loss": 1.2915,
+      "step": 23036
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8713568449020386,
+      "learning_rate": 4.391303198519693e-05,
+      "loss": 1.5605,
+      "step": 23037
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.5718116760253906,
+      "learning_rate": 4.391276764472641e-05,
+      "loss": 1.2931,
+      "step": 23038
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.5024499893188477,
+      "learning_rate": 4.3912503304255886e-05,
+      "loss": 1.3424,
+      "step": 23039
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0415427684783936,
+      "learning_rate": 4.391223896378536e-05,
+      "loss": 1.3411,
+      "step": 23040
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.949970006942749,
+      "learning_rate": 4.3911974623314833e-05,
+      "loss": 1.3818,
+      "step": 23041
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1209096908569336,
+      "learning_rate": 4.391171028284431e-05,
+      "loss": 1.3662,
+      "step": 23042
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1476995944976807,
+      "learning_rate": 4.391144594237378e-05,
+      "loss": 1.2589,
+      "step": 23043
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3990156650543213,
+      "learning_rate": 4.3911181601903254e-05,
+      "loss": 1.3887,
+      "step": 23044
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.213771343231201,
+      "learning_rate": 4.391091726143273e-05,
+      "loss": 1.212,
+      "step": 23045
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9341003894805908,
+      "learning_rate": 4.39106529209622e-05,
+      "loss": 1.1839,
+      "step": 23046
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7546818256378174,
+      "learning_rate": 4.3910388580491675e-05,
+      "loss": 1.2582,
+      "step": 23047
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.263760805130005,
+      "learning_rate": 4.391012424002115e-05,
+      "loss": 1.3508,
+      "step": 23048
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9518048763275146,
+      "learning_rate": 4.390985989955063e-05,
+      "loss": 1.2358,
+      "step": 23049
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.835951805114746,
+      "learning_rate": 4.3909595559080096e-05,
+      "loss": 1.3411,
+      "step": 23050
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.048159122467041,
+      "learning_rate": 4.3909331218609576e-05,
+      "loss": 1.4445,
+      "step": 23051
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8915081024169922,
+      "learning_rate": 4.390906687813904e-05,
+      "loss": 1.2594,
+      "step": 23052
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.5979379415512085,
+      "learning_rate": 4.3908802537668516e-05,
+      "loss": 1.2672,
+      "step": 23053
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.825659990310669,
+      "learning_rate": 4.3908538197198e-05,
+      "loss": 1.2639,
+      "step": 23054
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.053100109100342,
+      "learning_rate": 4.3908273856727463e-05,
+      "loss": 1.2359,
+      "step": 23055
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7406457662582397,
+      "learning_rate": 4.3908009516256944e-05,
+      "loss": 1.1408,
+      "step": 23056
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9497144222259521,
+      "learning_rate": 4.390774517578641e-05,
+      "loss": 1.2872,
+      "step": 23057
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7939975261688232,
+      "learning_rate": 4.390748083531589e-05,
+      "loss": 1.3398,
+      "step": 23058
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6968722343444824,
+      "learning_rate": 4.3907216494845365e-05,
+      "loss": 1.2266,
+      "step": 23059
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.5826280117034912,
+      "learning_rate": 4.390695215437484e-05,
+      "loss": 1.5664,
+      "step": 23060
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.722167491912842,
+      "learning_rate": 4.390668781390431e-05,
+      "loss": 1.3947,
+      "step": 23061
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.026068687438965,
+      "learning_rate": 4.3906423473433785e-05,
+      "loss": 1.2776,
+      "step": 23062
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.4773411750793457,
+      "learning_rate": 4.390615913296326e-05,
+      "loss": 1.1696,
+      "step": 23063
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.5412728786468506,
+      "learning_rate": 4.390589479249273e-05,
+      "loss": 1.4539,
+      "step": 23064
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2014739513397217,
+      "learning_rate": 4.3905630452022206e-05,
+      "loss": 1.1962,
+      "step": 23065
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.578317642211914,
+      "learning_rate": 4.390536611155168e-05,
+      "loss": 1.2381,
+      "step": 23066
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8545055389404297,
+      "learning_rate": 4.390510177108115e-05,
+      "loss": 1.2651,
+      "step": 23067
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3122682571411133,
+      "learning_rate": 4.390483743061063e-05,
+      "loss": 1.2958,
+      "step": 23068
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.916387915611267,
+      "learning_rate": 4.39045730901401e-05,
+      "loss": 1.2543,
+      "step": 23069
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3481483459472656,
+      "learning_rate": 4.3904308749669574e-05,
+      "loss": 1.424,
+      "step": 23070
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7129555940628052,
+      "learning_rate": 4.3904044409199054e-05,
+      "loss": 1.0597,
+      "step": 23071
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1193814277648926,
+      "learning_rate": 4.390378006872852e-05,
+      "loss": 1.4137,
+      "step": 23072
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9199849367141724,
+      "learning_rate": 4.3903515728258e-05,
+      "loss": 1.2673,
+      "step": 23073
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3069963455200195,
+      "learning_rate": 4.390325138778747e-05,
+      "loss": 1.3069,
+      "step": 23074
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7913250923156738,
+      "learning_rate": 4.390298704731695e-05,
+      "loss": 1.2359,
+      "step": 23075
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.354642152786255,
+      "learning_rate": 4.390272270684642e-05,
+      "loss": 1.6054,
+      "step": 23076
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8551620244979858,
+      "learning_rate": 4.3902458366375896e-05,
+      "loss": 1.2723,
+      "step": 23077
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9146186113357544,
+      "learning_rate": 4.390219402590537e-05,
+      "loss": 1.4232,
+      "step": 23078
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.724852442741394,
+      "learning_rate": 4.390192968543484e-05,
+      "loss": 2.2841,
+      "step": 23079
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.010577917098999,
+      "learning_rate": 4.3901665344964316e-05,
+      "loss": 1.1821,
+      "step": 23080
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.586062431335449,
+      "learning_rate": 4.390140100449379e-05,
+      "loss": 1.4589,
+      "step": 23081
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7999989986419678,
+      "learning_rate": 4.3901136664023263e-05,
+      "loss": 1.1992,
+      "step": 23082
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2747092247009277,
+      "learning_rate": 4.390087232355274e-05,
+      "loss": 1.3081,
+      "step": 23083
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8699151277542114,
+      "learning_rate": 4.390060798308221e-05,
+      "loss": 1.1457,
+      "step": 23084
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.5129631757736206,
+      "learning_rate": 4.3900343642611684e-05,
+      "loss": 1.8909,
+      "step": 23085
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.5016672611236572,
+      "learning_rate": 4.3900079302141165e-05,
+      "loss": 1.5652,
+      "step": 23086
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.208991765975952,
+      "learning_rate": 4.389981496167063e-05,
+      "loss": 1.4443,
+      "step": 23087
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.858961820602417,
+      "learning_rate": 4.389955062120011e-05,
+      "loss": 1.3368,
+      "step": 23088
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2661995887756348,
+      "learning_rate": 4.389928628072958e-05,
+      "loss": 1.314,
+      "step": 23089
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.567955255508423,
+      "learning_rate": 4.389902194025905e-05,
+      "loss": 1.4253,
+      "step": 23090
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1510138511657715,
+      "learning_rate": 4.389875759978853e-05,
+      "loss": 1.3685,
+      "step": 23091
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8235478401184082,
+      "learning_rate": 4.3898493259318e-05,
+      "loss": 1.3202,
+      "step": 23092
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2187297344207764,
+      "learning_rate": 4.389822891884748e-05,
+      "loss": 1.2428,
+      "step": 23093
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.4051315784454346,
+      "learning_rate": 4.3897964578376946e-05,
+      "loss": 1.3537,
+      "step": 23094
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.2306952476501465,
+      "learning_rate": 4.389770023790643e-05,
+      "loss": 1.3504,
+      "step": 23095
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.110074758529663,
+      "learning_rate": 4.38974358974359e-05,
+      "loss": 1.5175,
+      "step": 23096
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.7193403244018555,
+      "learning_rate": 4.3897171556965374e-05,
+      "loss": 1.5465,
+      "step": 23097
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.8024039268493652,
+      "learning_rate": 4.389690721649485e-05,
+      "loss": 1.5918,
+      "step": 23098
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.741494059562683,
+      "learning_rate": 4.389664287602432e-05,
+      "loss": 1.3235,
+      "step": 23099
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.145463466644287,
+      "learning_rate": 4.3896378535553795e-05,
+      "loss": 1.3455,
+      "step": 23100
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8259271383285522,
+      "learning_rate": 4.389611419508327e-05,
+      "loss": 1.2445,
+      "step": 23101
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.8847789764404297,
+      "learning_rate": 4.389584985461274e-05,
+      "loss": 1.3503,
+      "step": 23102
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.44140887260437,
+      "learning_rate": 4.3895585514142215e-05,
+      "loss": 1.3659,
+      "step": 23103
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9302234649658203,
+      "learning_rate": 4.389532117367169e-05,
+      "loss": 1.3678,
+      "step": 23104
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.7892849445343018,
+      "learning_rate": 4.389505683320116e-05,
+      "loss": 1.5001,
+      "step": 23105
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.820776343345642,
+      "learning_rate": 4.389479249273064e-05,
+      "loss": 1.1229,
+      "step": 23106
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2068774700164795,
+      "learning_rate": 4.389452815226011e-05,
+      "loss": 1.4391,
+      "step": 23107
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.8430795669555664,
+      "learning_rate": 4.389426381178959e-05,
+      "loss": 1.3735,
+      "step": 23108
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.051095485687256,
+      "learning_rate": 4.389399947131906e-05,
+      "loss": 1.2249,
+      "step": 23109
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7668100595474243,
+      "learning_rate": 4.389373513084854e-05,
+      "loss": 1.2209,
+      "step": 23110
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7779408693313599,
+      "learning_rate": 4.389347079037801e-05,
+      "loss": 1.3196,
+      "step": 23111
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7160921096801758,
+      "learning_rate": 4.3893206449907484e-05,
+      "loss": 1.2167,
+      "step": 23112
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3628575801849365,
+      "learning_rate": 4.389294210943696e-05,
+      "loss": 1.301,
+      "step": 23113
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2305893898010254,
+      "learning_rate": 4.389267776896643e-05,
+      "loss": 1.3241,
+      "step": 23114
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.037945508956909,
+      "learning_rate": 4.3892413428495905e-05,
+      "loss": 1.2973,
+      "step": 23115
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1100914478302,
+      "learning_rate": 4.389214908802538e-05,
+      "loss": 1.3109,
+      "step": 23116
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.95385479927063,
+      "learning_rate": 4.389188474755485e-05,
+      "loss": 1.4298,
+      "step": 23117
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7289183139801025,
+      "learning_rate": 4.3891620407084326e-05,
+      "loss": 1.2843,
+      "step": 23118
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.696296215057373,
+      "learning_rate": 4.38913560666138e-05,
+      "loss": 1.1494,
+      "step": 23119
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.972749948501587,
+      "learning_rate": 4.389109172614327e-05,
+      "loss": 1.3878,
+      "step": 23120
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8548496961593628,
+      "learning_rate": 4.389082738567275e-05,
+      "loss": 1.227,
+      "step": 23121
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1101088523864746,
+      "learning_rate": 4.389056304520222e-05,
+      "loss": 1.2882,
+      "step": 23122
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6818487644195557,
+      "learning_rate": 4.38902987047317e-05,
+      "loss": 1.3315,
+      "step": 23123
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.8344507217407227,
+      "learning_rate": 4.389003436426117e-05,
+      "loss": 1.6467,
+      "step": 23124
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.425443172454834,
+      "learning_rate": 4.388977002379065e-05,
+      "loss": 1.4489,
+      "step": 23125
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.226022481918335,
+      "learning_rate": 4.388950568332012e-05,
+      "loss": 1.3359,
+      "step": 23126
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.830004334449768,
+      "learning_rate": 4.388924134284959e-05,
+      "loss": 1.2037,
+      "step": 23127
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3588151931762695,
+      "learning_rate": 4.388897700237907e-05,
+      "loss": 1.3945,
+      "step": 23128
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.2710788249969482,
+      "learning_rate": 4.3888712661908535e-05,
+      "loss": 1.4047,
+      "step": 23129
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.6123242378234863,
+      "learning_rate": 4.3888448321438015e-05,
+      "loss": 1.4119,
+      "step": 23130
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9892979860305786,
+      "learning_rate": 4.388818398096749e-05,
+      "loss": 1.3431,
+      "step": 23131
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9575021266937256,
+      "learning_rate": 4.388791964049696e-05,
+      "loss": 1.3193,
+      "step": 23132
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0943851470947266,
+      "learning_rate": 4.3887655300026436e-05,
+      "loss": 1.3465,
+      "step": 23133
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.695204496383667,
+      "learning_rate": 4.388739095955591e-05,
+      "loss": 1.2438,
+      "step": 23134
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6759167909622192,
+      "learning_rate": 4.388712661908538e-05,
+      "loss": 3.3933,
+      "step": 23135
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.057025671005249,
+      "learning_rate": 4.388686227861486e-05,
+      "loss": 1.33,
+      "step": 23136
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7493886947631836,
+      "learning_rate": 4.388659793814433e-05,
+      "loss": 1.145,
+      "step": 23137
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.086669921875,
+      "learning_rate": 4.3886333597673804e-05,
+      "loss": 2.2898,
+      "step": 23138
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.494741916656494,
+      "learning_rate": 4.388606925720328e-05,
+      "loss": 1.2474,
+      "step": 23139
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7374416589736938,
+      "learning_rate": 4.388580491673275e-05,
+      "loss": 1.1602,
+      "step": 23140
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9560636281967163,
+      "learning_rate": 4.388554057626223e-05,
+      "loss": 1.2725,
+      "step": 23141
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.5662068128585815,
+      "learning_rate": 4.38852762357917e-05,
+      "loss": 1.0764,
+      "step": 23142
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8507822751998901,
+      "learning_rate": 4.388501189532118e-05,
+      "loss": 1.3094,
+      "step": 23143
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1792373657226562,
+      "learning_rate": 4.3884747554850645e-05,
+      "loss": 1.3902,
+      "step": 23144
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9245822429656982,
+      "learning_rate": 4.3884483214380126e-05,
+      "loss": 1.2522,
+      "step": 23145
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1560163497924805,
+      "learning_rate": 4.38842188739096e-05,
+      "loss": 1.4045,
+      "step": 23146
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0706429481506348,
+      "learning_rate": 4.388395453343907e-05,
+      "loss": 1.2232,
+      "step": 23147
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.5420260429382324,
+      "learning_rate": 4.3883690192968546e-05,
+      "loss": 1.2401,
+      "step": 23148
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.135456085205078,
+      "learning_rate": 4.388342585249802e-05,
+      "loss": 1.3506,
+      "step": 23149
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2569003105163574,
+      "learning_rate": 4.388316151202749e-05,
+      "loss": 1.2701,
+      "step": 23150
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.157022476196289,
+      "learning_rate": 4.388289717155697e-05,
+      "loss": 1.2637,
+      "step": 23151
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1916985511779785,
+      "learning_rate": 4.388263283108644e-05,
+      "loss": 1.2719,
+      "step": 23152
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0458385944366455,
+      "learning_rate": 4.3882368490615914e-05,
+      "loss": 1.3063,
+      "step": 23153
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.038522720336914,
+      "learning_rate": 4.388210415014539e-05,
+      "loss": 1.3083,
+      "step": 23154
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.916718602180481,
+      "learning_rate": 4.388183980967486e-05,
+      "loss": 1.3052,
+      "step": 23155
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.2375669479370117,
+      "learning_rate": 4.388157546920434e-05,
+      "loss": 1.3035,
+      "step": 23156
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.359116792678833,
+      "learning_rate": 4.388131112873381e-05,
+      "loss": 1.4162,
+      "step": 23157
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.252326488494873,
+      "learning_rate": 4.388104678826329e-05,
+      "loss": 1.3321,
+      "step": 23158
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0860040187835693,
+      "learning_rate": 4.3880782447792756e-05,
+      "loss": 1.2456,
+      "step": 23159
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.1402108669281006,
+      "learning_rate": 4.3880518107322236e-05,
+      "loss": 1.3113,
+      "step": 23160
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.166522741317749,
+      "learning_rate": 4.388025376685171e-05,
+      "loss": 1.3488,
+      "step": 23161
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9931920766830444,
+      "learning_rate": 4.3879989426381176e-05,
+      "loss": 1.4418,
+      "step": 23162
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.260479211807251,
+      "learning_rate": 4.3879725085910657e-05,
+      "loss": 1.3218,
+      "step": 23163
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8385963439941406,
+      "learning_rate": 4.387946074544012e-05,
+      "loss": 1.3701,
+      "step": 23164
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.357908010482788,
+      "learning_rate": 4.3879196404969604e-05,
+      "loss": 1.2933,
+      "step": 23165
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9596108198165894,
+      "learning_rate": 4.387893206449908e-05,
+      "loss": 1.3384,
+      "step": 23166
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7725441455841064,
+      "learning_rate": 4.387866772402855e-05,
+      "loss": 1.3733,
+      "step": 23167
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9553388357162476,
+      "learning_rate": 4.3878403383558024e-05,
+      "loss": 1.2965,
+      "step": 23168
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.6795899868011475,
+      "learning_rate": 4.38781390430875e-05,
+      "loss": 1.4101,
+      "step": 23169
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.596975326538086,
+      "learning_rate": 4.387787470261697e-05,
+      "loss": 1.1965,
+      "step": 23170
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.8028781414031982,
+      "learning_rate": 4.3877610362146445e-05,
+      "loss": 1.2168,
+      "step": 23171
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.5092403888702393,
+      "learning_rate": 4.387734602167592e-05,
+      "loss": 2.3894,
+      "step": 23172
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.789954662322998,
+      "learning_rate": 4.387708168120539e-05,
+      "loss": 2.203,
+      "step": 23173
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9328603744506836,
+      "learning_rate": 4.3876817340734866e-05,
+      "loss": 1.6378,
+      "step": 23174
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8919901847839355,
+      "learning_rate": 4.387655300026434e-05,
+      "loss": 1.231,
+      "step": 23175
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7576632499694824,
+      "learning_rate": 4.387628865979382e-05,
+      "loss": 1.1867,
+      "step": 23176
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.2318711280822754,
+      "learning_rate": 4.3876024319323287e-05,
+      "loss": 1.2092,
+      "step": 23177
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.604836344718933,
+      "learning_rate": 4.387575997885277e-05,
+      "loss": 1.396,
+      "step": 23178
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.223132371902466,
+      "learning_rate": 4.3875495638382234e-05,
+      "loss": 1.3043,
+      "step": 23179
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1218347549438477,
+      "learning_rate": 4.3875231297911714e-05,
+      "loss": 1.3196,
+      "step": 23180
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.582883834838867,
+      "learning_rate": 4.387496695744119e-05,
+      "loss": 1.4061,
+      "step": 23181
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8193793296813965,
+      "learning_rate": 4.387470261697066e-05,
+      "loss": 1.2939,
+      "step": 23182
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0511162281036377,
+      "learning_rate": 4.3874438276500135e-05,
+      "loss": 1.2563,
+      "step": 23183
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.993741512298584,
+      "learning_rate": 4.387417393602961e-05,
+      "loss": 1.3319,
+      "step": 23184
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1967434883117676,
+      "learning_rate": 4.387390959555908e-05,
+      "loss": 1.4293,
+      "step": 23185
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0554518699645996,
+      "learning_rate": 4.3873645255088555e-05,
+      "loss": 1.388,
+      "step": 23186
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7710320949554443,
+      "learning_rate": 4.387338091461803e-05,
+      "loss": 1.1378,
+      "step": 23187
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9145276546478271,
+      "learning_rate": 4.38731165741475e-05,
+      "loss": 1.435,
+      "step": 23188
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0768489837646484,
+      "learning_rate": 4.3872852233676976e-05,
+      "loss": 1.3614,
+      "step": 23189
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.244666576385498,
+      "learning_rate": 4.387258789320645e-05,
+      "loss": 1.3149,
+      "step": 23190
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8010362386703491,
+      "learning_rate": 4.387232355273593e-05,
+      "loss": 1.1392,
+      "step": 23191
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0105326175689697,
+      "learning_rate": 4.38720592122654e-05,
+      "loss": 1.2255,
+      "step": 23192
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.869143009185791,
+      "learning_rate": 4.387179487179488e-05,
+      "loss": 1.2423,
+      "step": 23193
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.808156132698059,
+      "learning_rate": 4.3871530531324344e-05,
+      "loss": 1.3046,
+      "step": 23194
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8751965761184692,
+      "learning_rate": 4.3871266190853824e-05,
+      "loss": 1.3628,
+      "step": 23195
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.283241033554077,
+      "learning_rate": 4.38710018503833e-05,
+      "loss": 1.3305,
+      "step": 23196
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6760197877883911,
+      "learning_rate": 4.387073750991277e-05,
+      "loss": 1.8548,
+      "step": 23197
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7952038049697876,
+      "learning_rate": 4.3870473169442245e-05,
+      "loss": 1.3314,
+      "step": 23198
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9113490581512451,
+      "learning_rate": 4.387020882897171e-05,
+      "loss": 1.2945,
+      "step": 23199
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.10229229927063,
+      "learning_rate": 4.386994448850119e-05,
+      "loss": 1.188,
+      "step": 23200
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1122422218322754,
+      "learning_rate": 4.3869680148030666e-05,
+      "loss": 1.256,
+      "step": 23201
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.2026138305664062,
+      "learning_rate": 4.386941580756014e-05,
+      "loss": 1.3678,
+      "step": 23202
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9407941102981567,
+      "learning_rate": 4.386915146708961e-05,
+      "loss": 1.2645,
+      "step": 23203
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.264840841293335,
+      "learning_rate": 4.3868887126619087e-05,
+      "loss": 1.2843,
+      "step": 23204
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.3455424308776855,
+      "learning_rate": 4.386862278614856e-05,
+      "loss": 1.3473,
+      "step": 23205
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0588955879211426,
+      "learning_rate": 4.386835844567804e-05,
+      "loss": 1.2901,
+      "step": 23206
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7945724725723267,
+      "learning_rate": 4.386809410520751e-05,
+      "loss": 1.5849,
+      "step": 23207
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.103921890258789,
+      "learning_rate": 4.386782976473698e-05,
+      "loss": 1.3986,
+      "step": 23208
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.014907121658325,
+      "learning_rate": 4.3867565424266454e-05,
+      "loss": 1.3793,
+      "step": 23209
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0379605293273926,
+      "learning_rate": 4.386730108379593e-05,
+      "loss": 1.8294,
+      "step": 23210
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.798701286315918,
+      "learning_rate": 4.386703674332541e-05,
+      "loss": 1.3157,
+      "step": 23211
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.5780627727508545,
+      "learning_rate": 4.3866772402854875e-05,
+      "loss": 1.3415,
+      "step": 23212
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8224912881851196,
+      "learning_rate": 4.3866508062384355e-05,
+      "loss": 1.2929,
+      "step": 23213
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7645349502563477,
+      "learning_rate": 4.386624372191382e-05,
+      "loss": 1.2401,
+      "step": 23214
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6165803670883179,
+      "learning_rate": 4.38659793814433e-05,
+      "loss": 1.6592,
+      "step": 23215
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.443052291870117,
+      "learning_rate": 4.3865715040972776e-05,
+      "loss": 1.3251,
+      "step": 23216
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.203238010406494,
+      "learning_rate": 4.386545070050225e-05,
+      "loss": 1.3415,
+      "step": 23217
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.801970362663269,
+      "learning_rate": 4.386518636003172e-05,
+      "loss": 1.1826,
+      "step": 23218
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0476889610290527,
+      "learning_rate": 4.38649220195612e-05,
+      "loss": 1.3287,
+      "step": 23219
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.406273603439331,
+      "learning_rate": 4.386465767909067e-05,
+      "loss": 1.3399,
+      "step": 23220
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0156421661376953,
+      "learning_rate": 4.3864393338620144e-05,
+      "loss": 1.3784,
+      "step": 23221
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.011631727218628,
+      "learning_rate": 4.386412899814962e-05,
+      "loss": 1.8806,
+      "step": 23222
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9085907936096191,
+      "learning_rate": 4.386386465767909e-05,
+      "loss": 1.2407,
+      "step": 23223
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0078320503234863,
+      "learning_rate": 4.3863600317208565e-05,
+      "loss": 1.537,
+      "step": 23224
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.990697979927063,
+      "learning_rate": 4.386333597673804e-05,
+      "loss": 1.2192,
+      "step": 23225
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8063088655471802,
+      "learning_rate": 4.386307163626752e-05,
+      "loss": 1.2417,
+      "step": 23226
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1822636127471924,
+      "learning_rate": 4.3862807295796985e-05,
+      "loss": 1.4261,
+      "step": 23227
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7942955493927002,
+      "learning_rate": 4.3862542955326466e-05,
+      "loss": 1.3363,
+      "step": 23228
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.654235601425171,
+      "learning_rate": 4.386227861485593e-05,
+      "loss": 1.3572,
+      "step": 23229
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.2755470275878906,
+      "learning_rate": 4.386201427438541e-05,
+      "loss": 1.4827,
+      "step": 23230
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.916531562805176,
+      "learning_rate": 4.3861749933914887e-05,
+      "loss": 1.5343,
+      "step": 23231
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.2161848545074463,
+      "learning_rate": 4.386148559344436e-05,
+      "loss": 1.4756,
+      "step": 23232
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9403761625289917,
+      "learning_rate": 4.3861221252973834e-05,
+      "loss": 1.4696,
+      "step": 23233
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.852179765701294,
+      "learning_rate": 4.38609569125033e-05,
+      "loss": 1.2376,
+      "step": 23234
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.050501585006714,
+      "learning_rate": 4.386069257203278e-05,
+      "loss": 1.2937,
+      "step": 23235
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1302733421325684,
+      "learning_rate": 4.3860428231562254e-05,
+      "loss": 1.4661,
+      "step": 23236
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.4311673641204834,
+      "learning_rate": 4.386016389109173e-05,
+      "loss": 1.4488,
+      "step": 23237
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0814173221588135,
+      "learning_rate": 4.38598995506212e-05,
+      "loss": 1.4038,
+      "step": 23238
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9751874208450317,
+      "learning_rate": 4.3859635210150675e-05,
+      "loss": 1.1825,
+      "step": 23239
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9655678272247314,
+      "learning_rate": 4.385937086968015e-05,
+      "loss": 1.2731,
+      "step": 23240
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.418034315109253,
+      "learning_rate": 4.385910652920963e-05,
+      "loss": 1.0715,
+      "step": 23241
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.8644025325775146,
+      "learning_rate": 4.3858842188739096e-05,
+      "loss": 1.2871,
+      "step": 23242
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6492916345596313,
+      "learning_rate": 4.385857784826857e-05,
+      "loss": 1.177,
+      "step": 23243
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.5097479820251465,
+      "learning_rate": 4.385831350779804e-05,
+      "loss": 1.39,
+      "step": 23244
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.174915313720703,
+      "learning_rate": 4.3858049167327517e-05,
+      "loss": 1.3589,
+      "step": 23245
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.872254729270935,
+      "learning_rate": 4.3857784826857e-05,
+      "loss": 1.3587,
+      "step": 23246
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.6127748489379883,
+      "learning_rate": 4.3857520486386464e-05,
+      "loss": 1.2951,
+      "step": 23247
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.885396122932434,
+      "learning_rate": 4.3857256145915944e-05,
+      "loss": 1.3373,
+      "step": 23248
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.05460786819458,
+      "learning_rate": 4.385699180544541e-05,
+      "loss": 1.2773,
+      "step": 23249
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0977234840393066,
+      "learning_rate": 4.385672746497489e-05,
+      "loss": 1.2131,
+      "step": 23250
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.2993321418762207,
+      "learning_rate": 4.3856463124504365e-05,
+      "loss": 1.4296,
+      "step": 23251
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9378918409347534,
+      "learning_rate": 4.385619878403384e-05,
+      "loss": 1.2802,
+      "step": 23252
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7256213426589966,
+      "learning_rate": 4.385593444356331e-05,
+      "loss": 1.4487,
+      "step": 23253
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.4423160552978516,
+      "learning_rate": 4.3855670103092785e-05,
+      "loss": 1.3526,
+      "step": 23254
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.20878005027771,
+      "learning_rate": 4.385540576262226e-05,
+      "loss": 1.4949,
+      "step": 23255
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.993584394454956,
+      "learning_rate": 4.385514142215173e-05,
+      "loss": 1.329,
+      "step": 23256
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.373642683029175,
+      "learning_rate": 4.3854877081681206e-05,
+      "loss": 1.3382,
+      "step": 23257
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1291444301605225,
+      "learning_rate": 4.385461274121068e-05,
+      "loss": 1.3808,
+      "step": 23258
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7913200855255127,
+      "learning_rate": 4.385434840074015e-05,
+      "loss": 1.2803,
+      "step": 23259
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7912760972976685,
+      "learning_rate": 4.385408406026963e-05,
+      "loss": 1.2181,
+      "step": 23260
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.8416543006896973,
+      "learning_rate": 4.385381971979911e-05,
+      "loss": 1.4369,
+      "step": 23261
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8184446096420288,
+      "learning_rate": 4.3853555379328574e-05,
+      "loss": 1.2858,
+      "step": 23262
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.713267207145691,
+      "learning_rate": 4.3853291038858054e-05,
+      "loss": 1.9567,
+      "step": 23263
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8600432872772217,
+      "learning_rate": 4.385302669838752e-05,
+      "loss": 1.2339,
+      "step": 23264
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0728721618652344,
+      "learning_rate": 4.3852762357917e-05,
+      "loss": 1.3765,
+      "step": 23265
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1698904037475586,
+      "learning_rate": 4.3852498017446475e-05,
+      "loss": 1.257,
+      "step": 23266
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8325258493423462,
+      "learning_rate": 4.385223367697595e-05,
+      "loss": 1.2919,
+      "step": 23267
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.321073293685913,
+      "learning_rate": 4.385196933650542e-05,
+      "loss": 1.2775,
+      "step": 23268
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9031745195388794,
+      "learning_rate": 4.3851704996034896e-05,
+      "loss": 1.2053,
+      "step": 23269
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.6179282665252686,
+      "learning_rate": 4.385144065556437e-05,
+      "loss": 1.3258,
+      "step": 23270
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6412577629089355,
+      "learning_rate": 4.385117631509384e-05,
+      "loss": 1.2123,
+      "step": 23271
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.11128306388855,
+      "learning_rate": 4.3850911974623316e-05,
+      "loss": 1.2138,
+      "step": 23272
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9436039924621582,
+      "learning_rate": 4.385064763415279e-05,
+      "loss": 1.2852,
+      "step": 23273
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.672732949256897,
+      "learning_rate": 4.3850383293682264e-05,
+      "loss": 1.2202,
+      "step": 23274
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9097111225128174,
+      "learning_rate": 4.385011895321174e-05,
+      "loss": 1.3156,
+      "step": 23275
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.119769334793091,
+      "learning_rate": 4.384985461274122e-05,
+      "loss": 1.3961,
+      "step": 23276
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.842828392982483,
+      "learning_rate": 4.3849590272270684e-05,
+      "loss": 1.3143,
+      "step": 23277
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.3777801990509033,
+      "learning_rate": 4.3849325931800165e-05,
+      "loss": 1.3898,
+      "step": 23278
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.092649221420288,
+      "learning_rate": 4.384906159132963e-05,
+      "loss": 1.311,
+      "step": 23279
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.5553979873657227,
+      "learning_rate": 4.3848797250859105e-05,
+      "loss": 1.4308,
+      "step": 23280
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.849847674369812,
+      "learning_rate": 4.3848532910388585e-05,
+      "loss": 1.2617,
+      "step": 23281
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1899917125701904,
+      "learning_rate": 4.384826856991805e-05,
+      "loss": 1.2508,
+      "step": 23282
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.1962740421295166,
+      "learning_rate": 4.384800422944753e-05,
+      "loss": 1.4936,
+      "step": 23283
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.2494330406188965,
+      "learning_rate": 4.3847739888977e-05,
+      "loss": 1.0998,
+      "step": 23284
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.32301926612854,
+      "learning_rate": 4.384747554850648e-05,
+      "loss": 1.1103,
+      "step": 23285
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0321428775787354,
+      "learning_rate": 4.384721120803595e-05,
+      "loss": 1.1958,
+      "step": 23286
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.204545736312866,
+      "learning_rate": 4.384694686756543e-05,
+      "loss": 1.2622,
+      "step": 23287
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6804566383361816,
+      "learning_rate": 4.38466825270949e-05,
+      "loss": 1.1318,
+      "step": 23288
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8993722200393677,
+      "learning_rate": 4.3846418186624374e-05,
+      "loss": 1.3083,
+      "step": 23289
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.7029764652252197,
+      "learning_rate": 4.384615384615385e-05,
+      "loss": 1.3873,
+      "step": 23290
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.289987087249756,
+      "learning_rate": 4.384588950568332e-05,
+      "loss": 1.3651,
+      "step": 23291
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8253121376037598,
+      "learning_rate": 4.3845625165212795e-05,
+      "loss": 1.254,
+      "step": 23292
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0279746055603027,
+      "learning_rate": 4.384536082474227e-05,
+      "loss": 1.2459,
+      "step": 23293
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.553348422050476,
+      "learning_rate": 4.384509648427174e-05,
+      "loss": 1.2498,
+      "step": 23294
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7194669246673584,
+      "learning_rate": 4.3844832143801215e-05,
+      "loss": 1.3643,
+      "step": 23295
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9661335945129395,
+      "learning_rate": 4.3844567803330696e-05,
+      "loss": 1.2435,
+      "step": 23296
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0965592861175537,
+      "learning_rate": 4.384430346286016e-05,
+      "loss": 1.448,
+      "step": 23297
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8022773265838623,
+      "learning_rate": 4.384403912238964e-05,
+      "loss": 1.159,
+      "step": 23298
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1256258487701416,
+      "learning_rate": 4.384377478191911e-05,
+      "loss": 1.3905,
+      "step": 23299
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.092315673828125,
+      "learning_rate": 4.384351044144859e-05,
+      "loss": 1.1131,
+      "step": 23300
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.6730291843414307,
+      "learning_rate": 4.3843246100978064e-05,
+      "loss": 1.3458,
+      "step": 23301
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1671369075775146,
+      "learning_rate": 4.384298176050754e-05,
+      "loss": 1.3385,
+      "step": 23302
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.382120370864868,
+      "learning_rate": 4.384271742003701e-05,
+      "loss": 1.3827,
+      "step": 23303
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.967418909072876,
+      "learning_rate": 4.3842453079566484e-05,
+      "loss": 1.1803,
+      "step": 23304
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7386683225631714,
+      "learning_rate": 4.384218873909596e-05,
+      "loss": 1.291,
+      "step": 23305
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.5911917686462402,
+      "learning_rate": 4.384192439862543e-05,
+      "loss": 1.2668,
+      "step": 23306
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0875585079193115,
+      "learning_rate": 4.3841660058154905e-05,
+      "loss": 1.3041,
+      "step": 23307
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.2476789951324463,
+      "learning_rate": 4.384139571768438e-05,
+      "loss": 1.4211,
+      "step": 23308
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6716482639312744,
+      "learning_rate": 4.384113137721385e-05,
+      "loss": 1.1205,
+      "step": 23309
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0115184783935547,
+      "learning_rate": 4.3840867036743326e-05,
+      "loss": 1.37,
+      "step": 23310
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.225120782852173,
+      "learning_rate": 4.3840602696272806e-05,
+      "loss": 1.3052,
+      "step": 23311
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.740159511566162,
+      "learning_rate": 4.384033835580227e-05,
+      "loss": 1.2428,
+      "step": 23312
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.364307165145874,
+      "learning_rate": 4.384007401533175e-05,
+      "loss": 1.467,
+      "step": 23313
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1375386714935303,
+      "learning_rate": 4.383980967486122e-05,
+      "loss": 1.3235,
+      "step": 23314
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8400999307632446,
+      "learning_rate": 4.38395453343907e-05,
+      "loss": 1.3049,
+      "step": 23315
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.465946674346924,
+      "learning_rate": 4.3839280993920174e-05,
+      "loss": 1.3347,
+      "step": 23316
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0792293548583984,
+      "learning_rate": 4.383901665344964e-05,
+      "loss": 1.4395,
+      "step": 23317
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.6088905334472656,
+      "learning_rate": 4.383875231297912e-05,
+      "loss": 1.4472,
+      "step": 23318
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0214684009552,
+      "learning_rate": 4.383848797250859e-05,
+      "loss": 1.3974,
+      "step": 23319
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6259065866470337,
+      "learning_rate": 4.383822363203807e-05,
+      "loss": 1.199,
+      "step": 23320
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.711662769317627,
+      "learning_rate": 4.383795929156754e-05,
+      "loss": 1.3725,
+      "step": 23321
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9117971658706665,
+      "learning_rate": 4.3837694951097015e-05,
+      "loss": 1.7278,
+      "step": 23322
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.423611640930176,
+      "learning_rate": 4.383743061062649e-05,
+      "loss": 1.4895,
+      "step": 23323
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.300504684448242,
+      "learning_rate": 4.383716627015596e-05,
+      "loss": 1.2889,
+      "step": 23324
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.816051721572876,
+      "learning_rate": 4.3836901929685436e-05,
+      "loss": 1.355,
+      "step": 23325
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.199009895324707,
+      "learning_rate": 4.383663758921491e-05,
+      "loss": 1.2989,
+      "step": 23326
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.3462626934051514,
+      "learning_rate": 4.383637324874438e-05,
+      "loss": 1.3134,
+      "step": 23327
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7399860620498657,
+      "learning_rate": 4.383610890827386e-05,
+      "loss": 1.3503,
+      "step": 23328
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9784777164459229,
+      "learning_rate": 4.383584456780333e-05,
+      "loss": 1.3143,
+      "step": 23329
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7716119289398193,
+      "learning_rate": 4.3835580227332804e-05,
+      "loss": 1.3206,
+      "step": 23330
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0784809589385986,
+      "learning_rate": 4.3835315886862284e-05,
+      "loss": 1.3715,
+      "step": 23331
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7612662315368652,
+      "learning_rate": 4.383505154639175e-05,
+      "loss": 1.0903,
+      "step": 23332
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0875344276428223,
+      "learning_rate": 4.383478720592123e-05,
+      "loss": 1.3396,
+      "step": 23333
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.4908788204193115,
+      "learning_rate": 4.38345228654507e-05,
+      "loss": 1.3207,
+      "step": 23334
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9391337633132935,
+      "learning_rate": 4.383425852498018e-05,
+      "loss": 1.3478,
+      "step": 23335
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.1231930255889893,
+      "learning_rate": 4.383399418450965e-05,
+      "loss": 1.4387,
+      "step": 23336
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0342774391174316,
+      "learning_rate": 4.3833729844039126e-05,
+      "loss": 1.1939,
+      "step": 23337
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7130769491195679,
+      "learning_rate": 4.38334655035686e-05,
+      "loss": 1.2728,
+      "step": 23338
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9089683294296265,
+      "learning_rate": 4.383320116309807e-05,
+      "loss": 1.2742,
+      "step": 23339
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.2031571865081787,
+      "learning_rate": 4.3832936822627546e-05,
+      "loss": 1.4394,
+      "step": 23340
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.683849334716797,
+      "learning_rate": 4.383267248215702e-05,
+      "loss": 1.4063,
+      "step": 23341
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.677394390106201,
+      "learning_rate": 4.3832408141686494e-05,
+      "loss": 1.486,
+      "step": 23342
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.4593212604522705,
+      "learning_rate": 4.383214380121597e-05,
+      "loss": 1.4835,
+      "step": 23343
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.915595293045044,
+      "learning_rate": 4.383187946074544e-05,
+      "loss": 1.2744,
+      "step": 23344
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.5224924087524414,
+      "learning_rate": 4.3831615120274914e-05,
+      "loss": 1.4352,
+      "step": 23345
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8765106201171875,
+      "learning_rate": 4.3831350779804395e-05,
+      "loss": 1.2103,
+      "step": 23346
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6407015323638916,
+      "learning_rate": 4.383108643933386e-05,
+      "loss": 1.1203,
+      "step": 23347
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.7498390674591064,
+      "learning_rate": 4.383082209886334e-05,
+      "loss": 1.3987,
+      "step": 23348
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.2557597160339355,
+      "learning_rate": 4.383055775839281e-05,
+      "loss": 1.2387,
+      "step": 23349
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.4436357021331787,
+      "learning_rate": 4.383029341792229e-05,
+      "loss": 1.3094,
+      "step": 23350
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.4956157207489014,
+      "learning_rate": 4.383002907745176e-05,
+      "loss": 1.3811,
+      "step": 23351
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.817731261253357,
+      "learning_rate": 4.382976473698123e-05,
+      "loss": 1.1385,
+      "step": 23352
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.2416515350341797,
+      "learning_rate": 4.382950039651071e-05,
+      "loss": 1.3952,
+      "step": 23353
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.158886194229126,
+      "learning_rate": 4.3829236056040176e-05,
+      "loss": 1.3907,
+      "step": 23354
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7942101955413818,
+      "learning_rate": 4.382897171556966e-05,
+      "loss": 1.2347,
+      "step": 23355
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8463047742843628,
+      "learning_rate": 4.382870737509913e-05,
+      "loss": 1.281,
+      "step": 23356
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.8290990591049194,
+      "learning_rate": 4.3828443034628604e-05,
+      "loss": 1.2782,
+      "step": 23357
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.352792263031006,
+      "learning_rate": 4.382817869415808e-05,
+      "loss": 1.2934,
+      "step": 23358
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.5668820142745972,
+      "learning_rate": 4.382791435368755e-05,
+      "loss": 2.0141,
+      "step": 23359
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.39748215675354,
+      "learning_rate": 4.3827650013217025e-05,
+      "loss": 1.3119,
+      "step": 23360
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.5348966121673584,
+      "learning_rate": 4.38273856727465e-05,
+      "loss": 1.2181,
+      "step": 23361
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.163778066635132,
+      "learning_rate": 4.382712133227597e-05,
+      "loss": 1.2896,
+      "step": 23362
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.680716633796692,
+      "learning_rate": 4.3826856991805445e-05,
+      "loss": 1.2679,
+      "step": 23363
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.14342999458313,
+      "learning_rate": 4.382659265133492e-05,
+      "loss": 1.4453,
+      "step": 23364
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.368746042251587,
+      "learning_rate": 4.382632831086439e-05,
+      "loss": 1.2857,
+      "step": 23365
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.6036250591278076,
+      "learning_rate": 4.382606397039387e-05,
+      "loss": 1.0744,
+      "step": 23366
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.967936396598816,
+      "learning_rate": 4.382579962992334e-05,
+      "loss": 1.2442,
+      "step": 23367
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2714662551879883,
+      "learning_rate": 4.382553528945282e-05,
+      "loss": 1.5092,
+      "step": 23368
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.958147644996643,
+      "learning_rate": 4.382527094898229e-05,
+      "loss": 1.1746,
+      "step": 23369
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9688777923583984,
+      "learning_rate": 4.382500660851177e-05,
+      "loss": 1.3147,
+      "step": 23370
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1943516731262207,
+      "learning_rate": 4.382474226804124e-05,
+      "loss": 1.3429,
+      "step": 23371
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9292471408843994,
+      "learning_rate": 4.3824477927570714e-05,
+      "loss": 1.3349,
+      "step": 23372
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.285285472869873,
+      "learning_rate": 4.382421358710019e-05,
+      "loss": 1.3296,
+      "step": 23373
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3073301315307617,
+      "learning_rate": 4.382394924662966e-05,
+      "loss": 1.4218,
+      "step": 23374
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.945110559463501,
+      "learning_rate": 4.3823684906159135e-05,
+      "loss": 1.2556,
+      "step": 23375
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8028796911239624,
+      "learning_rate": 4.382342056568861e-05,
+      "loss": 1.2043,
+      "step": 23376
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8502025604248047,
+      "learning_rate": 4.382315622521808e-05,
+      "loss": 1.2763,
+      "step": 23377
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6643290519714355,
+      "learning_rate": 4.3822891884747556e-05,
+      "loss": 1.2636,
+      "step": 23378
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9713574647903442,
+      "learning_rate": 4.382262754427703e-05,
+      "loss": 1.3223,
+      "step": 23379
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.632728099822998,
+      "learning_rate": 4.38223632038065e-05,
+      "loss": 2.0674,
+      "step": 23380
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.187401533126831,
+      "learning_rate": 4.382209886333598e-05,
+      "loss": 2.1186,
+      "step": 23381
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.0156242847442627,
+      "learning_rate": 4.382183452286545e-05,
+      "loss": 1.3083,
+      "step": 23382
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6124866008758545,
+      "learning_rate": 4.382157018239493e-05,
+      "loss": 1.2052,
+      "step": 23383
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.647797107696533,
+      "learning_rate": 4.38213058419244e-05,
+      "loss": 1.5332,
+      "step": 23384
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7973637580871582,
+      "learning_rate": 4.382104150145388e-05,
+      "loss": 1.1661,
+      "step": 23385
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.977837324142456,
+      "learning_rate": 4.382077716098335e-05,
+      "loss": 1.3283,
+      "step": 23386
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.427774429321289,
+      "learning_rate": 4.3820512820512825e-05,
+      "loss": 1.4603,
+      "step": 23387
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9757221937179565,
+      "learning_rate": 4.38202484800423e-05,
+      "loss": 1.397,
+      "step": 23388
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7210869789123535,
+      "learning_rate": 4.3819984139571765e-05,
+      "loss": 1.2408,
+      "step": 23389
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8870149850845337,
+      "learning_rate": 4.3819719799101245e-05,
+      "loss": 1.1644,
+      "step": 23390
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.4714511632919312,
+      "learning_rate": 4.381945545863072e-05,
+      "loss": 1.4919,
+      "step": 23391
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.5569020509719849,
+      "learning_rate": 4.381919111816019e-05,
+      "loss": 1.0738,
+      "step": 23392
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2255451679229736,
+      "learning_rate": 4.3818926777689666e-05,
+      "loss": 1.3631,
+      "step": 23393
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.9758687019348145,
+      "learning_rate": 4.381866243721914e-05,
+      "loss": 1.4199,
+      "step": 23394
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8401285409927368,
+      "learning_rate": 4.381839809674861e-05,
+      "loss": 1.1634,
+      "step": 23395
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9759198427200317,
+      "learning_rate": 4.3818133756278093e-05,
+      "loss": 1.3187,
+      "step": 23396
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.908665418624878,
+      "learning_rate": 4.381786941580756e-05,
+      "loss": 1.2556,
+      "step": 23397
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2580337524414062,
+      "learning_rate": 4.3817605075337034e-05,
+      "loss": 1.329,
+      "step": 23398
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9797215461730957,
+      "learning_rate": 4.381734073486651e-05,
+      "loss": 1.2417,
+      "step": 23399
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6868640184402466,
+      "learning_rate": 4.381707639439598e-05,
+      "loss": 1.1235,
+      "step": 23400
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7802191972732544,
+      "learning_rate": 4.381681205392546e-05,
+      "loss": 1.1897,
+      "step": 23401
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3546550273895264,
+      "learning_rate": 4.381654771345493e-05,
+      "loss": 1.9201,
+      "step": 23402
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0285017490386963,
+      "learning_rate": 4.381628337298441e-05,
+      "loss": 1.3609,
+      "step": 23403
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8674850463867188,
+      "learning_rate": 4.3816019032513875e-05,
+      "loss": 1.3373,
+      "step": 23404
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0358855724334717,
+      "learning_rate": 4.3815754692043356e-05,
+      "loss": 1.2461,
+      "step": 23405
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.576921224594116,
+      "learning_rate": 4.381549035157283e-05,
+      "loss": 1.2885,
+      "step": 23406
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9890613555908203,
+      "learning_rate": 4.38152260111023e-05,
+      "loss": 1.2857,
+      "step": 23407
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.7612102031707764,
+      "learning_rate": 4.3814961670631776e-05,
+      "loss": 1.5245,
+      "step": 23408
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6108912229537964,
+      "learning_rate": 4.381469733016125e-05,
+      "loss": 1.3386,
+      "step": 23409
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.04502010345459,
+      "learning_rate": 4.3814432989690723e-05,
+      "loss": 1.3927,
+      "step": 23410
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9852533340454102,
+      "learning_rate": 4.38141686492202e-05,
+      "loss": 1.1249,
+      "step": 23411
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9986740350723267,
+      "learning_rate": 4.381390430874967e-05,
+      "loss": 1.4883,
+      "step": 23412
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.772498607635498,
+      "learning_rate": 4.3813639968279144e-05,
+      "loss": 1.2539,
+      "step": 23413
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3116605281829834,
+      "learning_rate": 4.381337562780862e-05,
+      "loss": 1.393,
+      "step": 23414
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.152865171432495,
+      "learning_rate": 4.381311128733809e-05,
+      "loss": 1.3301,
+      "step": 23415
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.930730938911438,
+      "learning_rate": 4.381284694686757e-05,
+      "loss": 1.2035,
+      "step": 23416
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.4832106828689575,
+      "learning_rate": 4.381258260639704e-05,
+      "loss": 1.0221,
+      "step": 23417
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9916584491729736,
+      "learning_rate": 4.381231826592652e-05,
+      "loss": 2.1346,
+      "step": 23418
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6600689888000488,
+      "learning_rate": 4.3812053925455986e-05,
+      "loss": 1.2513,
+      "step": 23419
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9800745248794556,
+      "learning_rate": 4.3811789584985466e-05,
+      "loss": 1.2202,
+      "step": 23420
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3651273250579834,
+      "learning_rate": 4.381152524451494e-05,
+      "loss": 1.4782,
+      "step": 23421
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.53059458732605,
+      "learning_rate": 4.381126090404441e-05,
+      "loss": 1.4864,
+      "step": 23422
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8519439697265625,
+      "learning_rate": 4.381099656357389e-05,
+      "loss": 1.3549,
+      "step": 23423
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.193783760070801,
+      "learning_rate": 4.3810732223103353e-05,
+      "loss": 1.3479,
+      "step": 23424
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6953438520431519,
+      "learning_rate": 4.3810467882632834e-05,
+      "loss": 1.2474,
+      "step": 23425
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2376139163970947,
+      "learning_rate": 4.381020354216231e-05,
+      "loss": 1.4106,
+      "step": 23426
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1963233947753906,
+      "learning_rate": 4.380993920169178e-05,
+      "loss": 1.2735,
+      "step": 23427
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.613882303237915,
+      "learning_rate": 4.3809674861221255e-05,
+      "loss": 1.1043,
+      "step": 23428
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8767908811569214,
+      "learning_rate": 4.380941052075073e-05,
+      "loss": 1.2575,
+      "step": 23429
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.708621025085449,
+      "learning_rate": 4.38091461802802e-05,
+      "loss": 1.3605,
+      "step": 23430
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0540244579315186,
+      "learning_rate": 4.380888183980968e-05,
+      "loss": 1.4164,
+      "step": 23431
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7720658779144287,
+      "learning_rate": 4.380861749933915e-05,
+      "loss": 1.1456,
+      "step": 23432
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8730400800704956,
+      "learning_rate": 4.380835315886863e-05,
+      "loss": 1.7665,
+      "step": 23433
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1087961196899414,
+      "learning_rate": 4.3808088818398096e-05,
+      "loss": 1.2228,
+      "step": 23434
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.207029342651367,
+      "learning_rate": 4.380782447792757e-05,
+      "loss": 1.1937,
+      "step": 23435
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9840869903564453,
+      "learning_rate": 4.380756013745705e-05,
+      "loss": 1.4437,
+      "step": 23436
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.4984123706817627,
+      "learning_rate": 4.380729579698652e-05,
+      "loss": 1.4039,
+      "step": 23437
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.462419033050537,
+      "learning_rate": 4.3807031456516e-05,
+      "loss": 1.2538,
+      "step": 23438
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1073648929595947,
+      "learning_rate": 4.3806767116045464e-05,
+      "loss": 1.4832,
+      "step": 23439
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.061331033706665,
+      "learning_rate": 4.3806502775574944e-05,
+      "loss": 1.2971,
+      "step": 23440
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.760040283203125,
+      "learning_rate": 4.380623843510442e-05,
+      "loss": 1.2063,
+      "step": 23441
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.089465618133545,
+      "learning_rate": 4.380597409463389e-05,
+      "loss": 1.201,
+      "step": 23442
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.4668872356414795,
+      "learning_rate": 4.3805709754163365e-05,
+      "loss": 1.3687,
+      "step": 23443
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.976594090461731,
+      "learning_rate": 4.380544541369284e-05,
+      "loss": 1.2423,
+      "step": 23444
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.6188766956329346,
+      "learning_rate": 4.380518107322231e-05,
+      "loss": 1.5892,
+      "step": 23445
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3549258708953857,
+      "learning_rate": 4.3804916732751786e-05,
+      "loss": 1.2945,
+      "step": 23446
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9258676767349243,
+      "learning_rate": 4.380465239228126e-05,
+      "loss": 1.3713,
+      "step": 23447
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.6500704288482666,
+      "learning_rate": 4.380438805181073e-05,
+      "loss": 1.3046,
+      "step": 23448
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8812991380691528,
+      "learning_rate": 4.3804123711340206e-05,
+      "loss": 1.2904,
+      "step": 23449
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.08923077583313,
+      "learning_rate": 4.380385937086968e-05,
+      "loss": 1.2853,
+      "step": 23450
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7371317148208618,
+      "learning_rate": 4.380359503039916e-05,
+      "loss": 1.1994,
+      "step": 23451
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.424253225326538,
+      "learning_rate": 4.380333068992863e-05,
+      "loss": 1.3579,
+      "step": 23452
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.1638784408569336,
+      "learning_rate": 4.380306634945811e-05,
+      "loss": 1.3015,
+      "step": 23453
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6085968017578125,
+      "learning_rate": 4.3802802008987574e-05,
+      "loss": 1.2488,
+      "step": 23454
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9475910663604736,
+      "learning_rate": 4.3802537668517054e-05,
+      "loss": 1.2324,
+      "step": 23455
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7352218627929688,
+      "learning_rate": 4.380227332804653e-05,
+      "loss": 1.2814,
+      "step": 23456
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.322120428085327,
+      "learning_rate": 4.3802008987576e-05,
+      "loss": 1.423,
+      "step": 23457
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.977601170539856,
+      "learning_rate": 4.3801744647105475e-05,
+      "loss": 1.2581,
+      "step": 23458
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7618110179901123,
+      "learning_rate": 4.380148030663495e-05,
+      "loss": 1.2558,
+      "step": 23459
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.090162515640259,
+      "learning_rate": 4.380121596616442e-05,
+      "loss": 1.2332,
+      "step": 23460
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.4180550575256348,
+      "learning_rate": 4.3800951625693896e-05,
+      "loss": 1.1946,
+      "step": 23461
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3962960243225098,
+      "learning_rate": 4.380068728522337e-05,
+      "loss": 1.3575,
+      "step": 23462
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0513057708740234,
+      "learning_rate": 4.380042294475284e-05,
+      "loss": 1.3023,
+      "step": 23463
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.291163921356201,
+      "learning_rate": 4.380015860428232e-05,
+      "loss": 2.0239,
+      "step": 23464
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2014880180358887,
+      "learning_rate": 4.379989426381179e-05,
+      "loss": 1.6631,
+      "step": 23465
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.1655569076538086,
+      "learning_rate": 4.379962992334127e-05,
+      "loss": 1.4358,
+      "step": 23466
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.4923222064971924,
+      "learning_rate": 4.379936558287074e-05,
+      "loss": 1.2546,
+      "step": 23467
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.018371820449829,
+      "learning_rate": 4.379910124240022e-05,
+      "loss": 1.2897,
+      "step": 23468
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.5862958431243896,
+      "learning_rate": 4.3798836901929684e-05,
+      "loss": 1.3515,
+      "step": 23469
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0466768741607666,
+      "learning_rate": 4.379857256145916e-05,
+      "loss": 1.3684,
+      "step": 23470
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.861930012702942,
+      "learning_rate": 4.379830822098864e-05,
+      "loss": 1.4021,
+      "step": 23471
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8240950107574463,
+      "learning_rate": 4.3798043880518105e-05,
+      "loss": 1.3465,
+      "step": 23472
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3392062187194824,
+      "learning_rate": 4.3797779540047586e-05,
+      "loss": 1.5225,
+      "step": 23473
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.250750780105591,
+      "learning_rate": 4.379751519957705e-05,
+      "loss": 1.3593,
+      "step": 23474
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7039408683776855,
+      "learning_rate": 4.379725085910653e-05,
+      "loss": 1.1526,
+      "step": 23475
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.5974760055541992,
+      "learning_rate": 4.3796986518636006e-05,
+      "loss": 1.3108,
+      "step": 23476
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2348031997680664,
+      "learning_rate": 4.379672217816548e-05,
+      "loss": 1.3462,
+      "step": 23477
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.131427764892578,
+      "learning_rate": 4.3796457837694953e-05,
+      "loss": 1.4697,
+      "step": 23478
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.252687931060791,
+      "learning_rate": 4.379619349722443e-05,
+      "loss": 1.4046,
+      "step": 23479
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.121175527572632,
+      "learning_rate": 4.37959291567539e-05,
+      "loss": 1.2446,
+      "step": 23480
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2299957275390625,
+      "learning_rate": 4.3795664816283374e-05,
+      "loss": 1.3732,
+      "step": 23481
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6241105794906616,
+      "learning_rate": 4.379540047581285e-05,
+      "loss": 1.1476,
+      "step": 23482
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.5007612705230713,
+      "learning_rate": 4.379513613534232e-05,
+      "loss": 1.3774,
+      "step": 23483
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0803282260894775,
+      "learning_rate": 4.3794871794871795e-05,
+      "loss": 1.3722,
+      "step": 23484
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8825045824050903,
+      "learning_rate": 4.379460745440127e-05,
+      "loss": 1.198,
+      "step": 23485
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8956801891326904,
+      "learning_rate": 4.379434311393075e-05,
+      "loss": 1.2984,
+      "step": 23486
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8859492540359497,
+      "learning_rate": 4.3794078773460216e-05,
+      "loss": 1.2657,
+      "step": 23487
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.024055004119873,
+      "learning_rate": 4.3793814432989696e-05,
+      "loss": 1.2452,
+      "step": 23488
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7102547883987427,
+      "learning_rate": 4.379355009251916e-05,
+      "loss": 1.2979,
+      "step": 23489
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.354466676712036,
+      "learning_rate": 4.379328575204864e-05,
+      "loss": 1.3794,
+      "step": 23490
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8676578998565674,
+      "learning_rate": 4.3793021411578117e-05,
+      "loss": 1.1993,
+      "step": 23491
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.501028537750244,
+      "learning_rate": 4.379275707110759e-05,
+      "loss": 1.3574,
+      "step": 23492
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3063113689422607,
+      "learning_rate": 4.3792492730637064e-05,
+      "loss": 1.3198,
+      "step": 23493
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.598413109779358,
+      "learning_rate": 4.379222839016654e-05,
+      "loss": 1.298,
+      "step": 23494
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.927028775215149,
+      "learning_rate": 4.379196404969601e-05,
+      "loss": 1.6072,
+      "step": 23495
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.547009229660034,
+      "learning_rate": 4.3791699709225484e-05,
+      "loss": 1.5043,
+      "step": 23496
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.769564151763916,
+      "learning_rate": 4.379143536875496e-05,
+      "loss": 1.4034,
+      "step": 23497
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.028672933578491,
+      "learning_rate": 4.379117102828443e-05,
+      "loss": 1.3915,
+      "step": 23498
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.7024450302124023,
+      "learning_rate": 4.3790906687813905e-05,
+      "loss": 1.5247,
+      "step": 23499
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.148538589477539,
+      "learning_rate": 4.379064234734338e-05,
+      "loss": 1.3033,
+      "step": 23500
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3683931827545166,
+      "learning_rate": 4.379037800687286e-05,
+      "loss": 1.3696,
+      "step": 23501
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1446731090545654,
+      "learning_rate": 4.3790113666402326e-05,
+      "loss": 1.4803,
+      "step": 23502
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7886244058609009,
+      "learning_rate": 4.3789849325931806e-05,
+      "loss": 1.2527,
+      "step": 23503
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8992924690246582,
+      "learning_rate": 4.378958498546127e-05,
+      "loss": 1.3842,
+      "step": 23504
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7969259023666382,
+      "learning_rate": 4.378932064499075e-05,
+      "loss": 1.2686,
+      "step": 23505
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.26010799407959,
+      "learning_rate": 4.378905630452023e-05,
+      "loss": 1.4045,
+      "step": 23506
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.34714937210083,
+      "learning_rate": 4.3788791964049694e-05,
+      "loss": 1.2821,
+      "step": 23507
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7404978275299072,
+      "learning_rate": 4.3788527623579174e-05,
+      "loss": 1.2202,
+      "step": 23508
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.82561457157135,
+      "learning_rate": 4.378826328310864e-05,
+      "loss": 1.1276,
+      "step": 23509
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9074082374572754,
+      "learning_rate": 4.378799894263812e-05,
+      "loss": 1.0655,
+      "step": 23510
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9568743705749512,
+      "learning_rate": 4.3787734602167595e-05,
+      "loss": 1.3495,
+      "step": 23511
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7552194595336914,
+      "learning_rate": 4.378747026169707e-05,
+      "loss": 1.2316,
+      "step": 23512
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1816577911376953,
+      "learning_rate": 4.378720592122654e-05,
+      "loss": 1.3082,
+      "step": 23513
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2546043395996094,
+      "learning_rate": 4.3786941580756016e-05,
+      "loss": 1.4921,
+      "step": 23514
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8499462604522705,
+      "learning_rate": 4.378667724028549e-05,
+      "loss": 1.2257,
+      "step": 23515
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1655666828155518,
+      "learning_rate": 4.378641289981496e-05,
+      "loss": 1.2564,
+      "step": 23516
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.849990725517273,
+      "learning_rate": 4.3786148559344436e-05,
+      "loss": 1.2361,
+      "step": 23517
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.3804962635040283,
+      "learning_rate": 4.378588421887391e-05,
+      "loss": 1.1058,
+      "step": 23518
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.871978521347046,
+      "learning_rate": 4.378561987840338e-05,
+      "loss": 1.2389,
+      "step": 23519
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.091515064239502,
+      "learning_rate": 4.378535553793286e-05,
+      "loss": 1.3746,
+      "step": 23520
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.229503631591797,
+      "learning_rate": 4.378509119746234e-05,
+      "loss": 1.2641,
+      "step": 23521
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.6787850856781006,
+      "learning_rate": 4.3784826856991804e-05,
+      "loss": 1.3619,
+      "step": 23522
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.389078378677368,
+      "learning_rate": 4.3784562516521284e-05,
+      "loss": 1.186,
+      "step": 23523
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.052264451980591,
+      "learning_rate": 4.378429817605075e-05,
+      "loss": 1.2913,
+      "step": 23524
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.3636038303375244,
+      "learning_rate": 4.378403383558023e-05,
+      "loss": 1.2053,
+      "step": 23525
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.711747169494629,
+      "learning_rate": 4.3783769495109705e-05,
+      "loss": 1.1451,
+      "step": 23526
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.7219696044921875,
+      "learning_rate": 4.378350515463918e-05,
+      "loss": 1.4272,
+      "step": 23527
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.126584768295288,
+      "learning_rate": 4.378324081416865e-05,
+      "loss": 1.3152,
+      "step": 23528
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.131387710571289,
+      "learning_rate": 4.3782976473698126e-05,
+      "loss": 1.1988,
+      "step": 23529
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.459686756134033,
+      "learning_rate": 4.37827121332276e-05,
+      "loss": 1.3426,
+      "step": 23530
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.807997226715088,
+      "learning_rate": 4.378244779275707e-05,
+      "loss": 1.1794,
+      "step": 23531
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.491732120513916,
+      "learning_rate": 4.3782183452286547e-05,
+      "loss": 1.2933,
+      "step": 23532
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6075114011764526,
+      "learning_rate": 4.378191911181602e-05,
+      "loss": 1.253,
+      "step": 23533
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.454915761947632,
+      "learning_rate": 4.3781654771345494e-05,
+      "loss": 1.2819,
+      "step": 23534
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6258766651153564,
+      "learning_rate": 4.378139043087497e-05,
+      "loss": 1.248,
+      "step": 23535
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.317891836166382,
+      "learning_rate": 4.378112609040445e-05,
+      "loss": 1.1625,
+      "step": 23536
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.6312196254730225,
+      "learning_rate": 4.3780861749933914e-05,
+      "loss": 1.432,
+      "step": 23537
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.379297971725464,
+      "learning_rate": 4.3780597409463395e-05,
+      "loss": 1.4015,
+      "step": 23538
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.741307258605957,
+      "learning_rate": 4.378033306899286e-05,
+      "loss": 1.3088,
+      "step": 23539
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.062727928161621,
+      "learning_rate": 4.378006872852234e-05,
+      "loss": 1.3812,
+      "step": 23540
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.383574962615967,
+      "learning_rate": 4.3779804388051815e-05,
+      "loss": 1.3498,
+      "step": 23541
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0200936794281006,
+      "learning_rate": 4.377954004758128e-05,
+      "loss": 1.4379,
+      "step": 23542
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2414767742156982,
+      "learning_rate": 4.377927570711076e-05,
+      "loss": 1.1351,
+      "step": 23543
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8789106607437134,
+      "learning_rate": 4.377901136664023e-05,
+      "loss": 1.2012,
+      "step": 23544
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.067065715789795,
+      "learning_rate": 4.377874702616971e-05,
+      "loss": 1.3487,
+      "step": 23545
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9091062545776367,
+      "learning_rate": 4.377848268569918e-05,
+      "loss": 1.2001,
+      "step": 23546
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.8890087604522705,
+      "learning_rate": 4.377821834522866e-05,
+      "loss": 2.0437,
+      "step": 23547
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9846699237823486,
+      "learning_rate": 4.377795400475813e-05,
+      "loss": 1.1947,
+      "step": 23548
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.229949474334717,
+      "learning_rate": 4.3777689664287604e-05,
+      "loss": 1.4094,
+      "step": 23549
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.3388547897338867,
+      "learning_rate": 4.377742532381708e-05,
+      "loss": 1.4928,
+      "step": 23550
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0951223373413086,
+      "learning_rate": 4.377716098334656e-05,
+      "loss": 1.2984,
+      "step": 23551
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.289644718170166,
+      "learning_rate": 4.3776896642876025e-05,
+      "loss": 1.3465,
+      "step": 23552
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.068091630935669,
+      "learning_rate": 4.37766323024055e-05,
+      "loss": 1.3992,
+      "step": 23553
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8610972166061401,
+      "learning_rate": 4.377636796193497e-05,
+      "loss": 1.2241,
+      "step": 23554
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.3118956089019775,
+      "learning_rate": 4.3776103621464445e-05,
+      "loss": 1.4249,
+      "step": 23555
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.4289910793304443,
+      "learning_rate": 4.3775839280993926e-05,
+      "loss": 1.3427,
+      "step": 23556
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.249004602432251,
+      "learning_rate": 4.377557494052339e-05,
+      "loss": 1.178,
+      "step": 23557
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9361135959625244,
+      "learning_rate": 4.377531060005287e-05,
+      "loss": 1.312,
+      "step": 23558
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.792513132095337,
+      "learning_rate": 4.377504625958234e-05,
+      "loss": 1.2749,
+      "step": 23559
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.621821403503418,
+      "learning_rate": 4.377478191911182e-05,
+      "loss": 1.5119,
+      "step": 23560
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.970428466796875,
+      "learning_rate": 4.3774517578641294e-05,
+      "loss": 1.3484,
+      "step": 23561
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.345987319946289,
+      "learning_rate": 4.377425323817077e-05,
+      "loss": 1.3409,
+      "step": 23562
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7509897947311401,
+      "learning_rate": 4.377398889770024e-05,
+      "loss": 1.2196,
+      "step": 23563
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8265892267227173,
+      "learning_rate": 4.3773724557229714e-05,
+      "loss": 2.0601,
+      "step": 23564
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.693636178970337,
+      "learning_rate": 4.377346021675919e-05,
+      "loss": 1.2147,
+      "step": 23565
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.835909366607666,
+      "learning_rate": 4.377319587628866e-05,
+      "loss": 1.2622,
+      "step": 23566
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1476492881774902,
+      "learning_rate": 4.3772931535818135e-05,
+      "loss": 1.2282,
+      "step": 23567
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0396058559417725,
+      "learning_rate": 4.377266719534761e-05,
+      "loss": 1.1886,
+      "step": 23568
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6799399852752686,
+      "learning_rate": 4.377240285487708e-05,
+      "loss": 1.219,
+      "step": 23569
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.4673826694488525,
+      "learning_rate": 4.3772138514406556e-05,
+      "loss": 1.3378,
+      "step": 23570
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.8106884956359863,
+      "learning_rate": 4.3771874173936036e-05,
+      "loss": 1.3887,
+      "step": 23571
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0071518421173096,
+      "learning_rate": 4.37716098334655e-05,
+      "loss": 1.2298,
+      "step": 23572
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8852405548095703,
+      "learning_rate": 4.377134549299498e-05,
+      "loss": 1.2695,
+      "step": 23573
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0459144115448,
+      "learning_rate": 4.377108115252445e-05,
+      "loss": 1.2732,
+      "step": 23574
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1965954303741455,
+      "learning_rate": 4.377081681205393e-05,
+      "loss": 1.3304,
+      "step": 23575
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0036067962646484,
+      "learning_rate": 4.3770552471583404e-05,
+      "loss": 1.1425,
+      "step": 23576
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1042683124542236,
+      "learning_rate": 4.377028813111288e-05,
+      "loss": 1.382,
+      "step": 23577
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0309972763061523,
+      "learning_rate": 4.377002379064235e-05,
+      "loss": 1.2262,
+      "step": 23578
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.230553150177002,
+      "learning_rate": 4.376975945017182e-05,
+      "loss": 1.281,
+      "step": 23579
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9990959167480469,
+      "learning_rate": 4.37694951097013e-05,
+      "loss": 1.3933,
+      "step": 23580
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0428407192230225,
+      "learning_rate": 4.376923076923077e-05,
+      "loss": 1.1504,
+      "step": 23581
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.7023019790649414,
+      "learning_rate": 4.3768966428760245e-05,
+      "loss": 1.5012,
+      "step": 23582
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1952733993530273,
+      "learning_rate": 4.376870208828972e-05,
+      "loss": 1.2451,
+      "step": 23583
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1220755577087402,
+      "learning_rate": 4.376843774781919e-05,
+      "loss": 1.661,
+      "step": 23584
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.994372844696045,
+      "learning_rate": 4.3768173407348666e-05,
+      "loss": 1.3352,
+      "step": 23585
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.5025806427001953,
+      "learning_rate": 4.3767909066878146e-05,
+      "loss": 1.4021,
+      "step": 23586
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.008727788925171,
+      "learning_rate": 4.376764472640761e-05,
+      "loss": 1.2581,
+      "step": 23587
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0915536880493164,
+      "learning_rate": 4.376738038593709e-05,
+      "loss": 1.2768,
+      "step": 23588
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7918922901153564,
+      "learning_rate": 4.376711604546656e-05,
+      "loss": 1.2083,
+      "step": 23589
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0259461402893066,
+      "learning_rate": 4.3766851704996034e-05,
+      "loss": 1.2406,
+      "step": 23590
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9260035753250122,
+      "learning_rate": 4.3766587364525514e-05,
+      "loss": 1.4522,
+      "step": 23591
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8046770095825195,
+      "learning_rate": 4.376632302405498e-05,
+      "loss": 1.3357,
+      "step": 23592
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.062169075012207,
+      "learning_rate": 4.376605868358446e-05,
+      "loss": 1.4082,
+      "step": 23593
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7709527015686035,
+      "learning_rate": 4.376579434311393e-05,
+      "loss": 1.2791,
+      "step": 23594
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7691317796707153,
+      "learning_rate": 4.376553000264341e-05,
+      "loss": 1.1929,
+      "step": 23595
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.275301456451416,
+      "learning_rate": 4.376526566217288e-05,
+      "loss": 1.4473,
+      "step": 23596
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1317546367645264,
+      "learning_rate": 4.3765001321702356e-05,
+      "loss": 1.3599,
+      "step": 23597
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9205282926559448,
+      "learning_rate": 4.376473698123183e-05,
+      "loss": 1.2266,
+      "step": 23598
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9148313999176025,
+      "learning_rate": 4.37644726407613e-05,
+      "loss": 1.3441,
+      "step": 23599
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.863383412361145,
+      "learning_rate": 4.3764208300290776e-05,
+      "loss": 1.0652,
+      "step": 23600
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0535683631896973,
+      "learning_rate": 4.376394395982025e-05,
+      "loss": 2.256,
+      "step": 23601
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1582796573638916,
+      "learning_rate": 4.3763679619349724e-05,
+      "loss": 1.3971,
+      "step": 23602
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.2449474334716797,
+      "learning_rate": 4.37634152788792e-05,
+      "loss": 1.3365,
+      "step": 23603
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7328404188156128,
+      "learning_rate": 4.376315093840867e-05,
+      "loss": 1.3274,
+      "step": 23604
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8650530576705933,
+      "learning_rate": 4.3762886597938144e-05,
+      "loss": 1.365,
+      "step": 23605
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4505934715270996,
+      "learning_rate": 4.3762622257467625e-05,
+      "loss": 1.1424,
+      "step": 23606
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.364729166030884,
+      "learning_rate": 4.376235791699709e-05,
+      "loss": 1.4948,
+      "step": 23607
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9835736751556396,
+      "learning_rate": 4.376209357652657e-05,
+      "loss": 1.3926,
+      "step": 23608
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8805770874023438,
+      "learning_rate": 4.376182923605604e-05,
+      "loss": 1.381,
+      "step": 23609
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8798819780349731,
+      "learning_rate": 4.376156489558552e-05,
+      "loss": 1.3012,
+      "step": 23610
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1064059734344482,
+      "learning_rate": 4.376130055511499e-05,
+      "loss": 1.3537,
+      "step": 23611
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8699926137924194,
+      "learning_rate": 4.3761036214644466e-05,
+      "loss": 1.302,
+      "step": 23612
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0573511123657227,
+      "learning_rate": 4.376077187417394e-05,
+      "loss": 1.4643,
+      "step": 23613
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.841641902923584,
+      "learning_rate": 4.3760507533703406e-05,
+      "loss": 1.1128,
+      "step": 23614
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.9523000717163086,
+      "learning_rate": 4.376024319323289e-05,
+      "loss": 1.288,
+      "step": 23615
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.5741667747497559,
+      "learning_rate": 4.375997885276236e-05,
+      "loss": 1.2424,
+      "step": 23616
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8612478971481323,
+      "learning_rate": 4.3759714512291834e-05,
+      "loss": 1.3656,
+      "step": 23617
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8413081169128418,
+      "learning_rate": 4.375945017182131e-05,
+      "loss": 1.2699,
+      "step": 23618
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.140296459197998,
+      "learning_rate": 4.375918583135078e-05,
+      "loss": 1.5976,
+      "step": 23619
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.854299545288086,
+      "learning_rate": 4.3758921490880255e-05,
+      "loss": 1.425,
+      "step": 23620
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.3263888359069824,
+      "learning_rate": 4.3758657150409735e-05,
+      "loss": 1.251,
+      "step": 23621
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7940458059310913,
+      "learning_rate": 4.37583928099392e-05,
+      "loss": 1.1817,
+      "step": 23622
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.82378888130188,
+      "learning_rate": 4.375812846946868e-05,
+      "loss": 1.4901,
+      "step": 23623
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9917269945144653,
+      "learning_rate": 4.375786412899815e-05,
+      "loss": 1.1424,
+      "step": 23624
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0690195560455322,
+      "learning_rate": 4.375759978852762e-05,
+      "loss": 1.3558,
+      "step": 23625
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.3119401931762695,
+      "learning_rate": 4.37573354480571e-05,
+      "loss": 1.3203,
+      "step": 23626
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.8471267223358154,
+      "learning_rate": 4.375707110758657e-05,
+      "loss": 1.433,
+      "step": 23627
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.574007034301758,
+      "learning_rate": 4.375680676711605e-05,
+      "loss": 1.3135,
+      "step": 23628
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.97510826587677,
+      "learning_rate": 4.375654242664552e-05,
+      "loss": 1.2363,
+      "step": 23629
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6943384408950806,
+      "learning_rate": 4.3756278086175e-05,
+      "loss": 1.1997,
+      "step": 23630
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.2454237937927246,
+      "learning_rate": 4.375601374570447e-05,
+      "loss": 1.2903,
+      "step": 23631
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1575725078582764,
+      "learning_rate": 4.3755749405233944e-05,
+      "loss": 1.2901,
+      "step": 23632
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8205373287200928,
+      "learning_rate": 4.375548506476342e-05,
+      "loss": 1.0576,
+      "step": 23633
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.6200287342071533,
+      "learning_rate": 4.375522072429289e-05,
+      "loss": 1.3622,
+      "step": 23634
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.4763288497924805,
+      "learning_rate": 4.3754956383822365e-05,
+      "loss": 1.5453,
+      "step": 23635
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.053179979324341,
+      "learning_rate": 4.375469204335184e-05,
+      "loss": 1.2682,
+      "step": 23636
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.388110876083374,
+      "learning_rate": 4.375442770288131e-05,
+      "loss": 1.2753,
+      "step": 23637
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0746829509735107,
+      "learning_rate": 4.3754163362410786e-05,
+      "loss": 1.4357,
+      "step": 23638
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.3550846576690674,
+      "learning_rate": 4.375389902194026e-05,
+      "loss": 1.4467,
+      "step": 23639
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1569433212280273,
+      "learning_rate": 4.375363468146973e-05,
+      "loss": 1.2824,
+      "step": 23640
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.673746943473816,
+      "learning_rate": 4.375337034099921e-05,
+      "loss": 1.051,
+      "step": 23641
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.2096595764160156,
+      "learning_rate": 4.375310600052868e-05,
+      "loss": 1.784,
+      "step": 23642
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6330913305282593,
+      "learning_rate": 4.375284166005816e-05,
+      "loss": 1.2402,
+      "step": 23643
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.177657127380371,
+      "learning_rate": 4.375257731958763e-05,
+      "loss": 1.3253,
+      "step": 23644
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.3611223697662354,
+      "learning_rate": 4.375231297911711e-05,
+      "loss": 1.3427,
+      "step": 23645
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.374183177947998,
+      "learning_rate": 4.375204863864658e-05,
+      "loss": 1.4286,
+      "step": 23646
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.7742912769317627,
+      "learning_rate": 4.3751784298176055e-05,
+      "loss": 1.3575,
+      "step": 23647
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.3767313957214355,
+      "learning_rate": 4.375151995770553e-05,
+      "loss": 1.3833,
+      "step": 23648
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.2627243995666504,
+      "learning_rate": 4.3751255617235e-05,
+      "loss": 1.4002,
+      "step": 23649
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.2378363609313965,
+      "learning_rate": 4.3750991276764475e-05,
+      "loss": 1.2858,
+      "step": 23650
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.186635732650757,
+      "learning_rate": 4.375072693629395e-05,
+      "loss": 1.2444,
+      "step": 23651
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0292727947235107,
+      "learning_rate": 4.375046259582342e-05,
+      "loss": 1.3861,
+      "step": 23652
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.024254322052002,
+      "learning_rate": 4.3750198255352896e-05,
+      "loss": 1.3115,
+      "step": 23653
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.032306432723999,
+      "learning_rate": 4.374993391488237e-05,
+      "loss": 1.4421,
+      "step": 23654
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1177191734313965,
+      "learning_rate": 4.374966957441184e-05,
+      "loss": 1.2827,
+      "step": 23655
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.234135866165161,
+      "learning_rate": 4.3749405233941324e-05,
+      "loss": 1.4091,
+      "step": 23656
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9200160503387451,
+      "learning_rate": 4.374914089347079e-05,
+      "loss": 1.265,
+      "step": 23657
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.294854164123535,
+      "learning_rate": 4.374887655300027e-05,
+      "loss": 1.3358,
+      "step": 23658
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9869784116744995,
+      "learning_rate": 4.374861221252974e-05,
+      "loss": 1.3737,
+      "step": 23659
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.2670705318450928,
+      "learning_rate": 4.374834787205921e-05,
+      "loss": 1.3424,
+      "step": 23660
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7694870233535767,
+      "learning_rate": 4.374808353158869e-05,
+      "loss": 1.1407,
+      "step": 23661
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1765193939208984,
+      "learning_rate": 4.374781919111816e-05,
+      "loss": 1.3076,
+      "step": 23662
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6896569728851318,
+      "learning_rate": 4.374755485064764e-05,
+      "loss": 1.3549,
+      "step": 23663
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1263556480407715,
+      "learning_rate": 4.3747290510177105e-05,
+      "loss": 1.1715,
+      "step": 23664
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.6578216552734375,
+      "learning_rate": 4.3747026169706586e-05,
+      "loss": 1.6147,
+      "step": 23665
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.3506083488464355,
+      "learning_rate": 4.374676182923606e-05,
+      "loss": 1.1851,
+      "step": 23666
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8355308771133423,
+      "learning_rate": 4.374649748876553e-05,
+      "loss": 1.1309,
+      "step": 23667
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.060976028442383,
+      "learning_rate": 4.3746233148295006e-05,
+      "loss": 1.1641,
+      "step": 23668
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.439835548400879,
+      "learning_rate": 4.374596880782448e-05,
+      "loss": 1.3977,
+      "step": 23669
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1741974353790283,
+      "learning_rate": 4.3745704467353954e-05,
+      "loss": 1.3616,
+      "step": 23670
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.849279761314392,
+      "learning_rate": 4.374544012688343e-05,
+      "loss": 1.3341,
+      "step": 23671
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.639306902885437,
+      "learning_rate": 4.37451757864129e-05,
+      "loss": 1.2852,
+      "step": 23672
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.5171642303466797,
+      "learning_rate": 4.3744911445942374e-05,
+      "loss": 1.5715,
+      "step": 23673
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.5912834405899048,
+      "learning_rate": 4.374464710547185e-05,
+      "loss": 1.1658,
+      "step": 23674
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.3913676738739014,
+      "learning_rate": 4.374438276500132e-05,
+      "loss": 1.3979,
+      "step": 23675
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7632900476455688,
+      "learning_rate": 4.37441184245308e-05,
+      "loss": 1.0563,
+      "step": 23676
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6966439485549927,
+      "learning_rate": 4.374385408406027e-05,
+      "loss": 1.2952,
+      "step": 23677
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7035586833953857,
+      "learning_rate": 4.374358974358975e-05,
+      "loss": 1.0656,
+      "step": 23678
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8507823944091797,
+      "learning_rate": 4.3743325403119216e-05,
+      "loss": 1.2167,
+      "step": 23679
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.002484083175659,
+      "learning_rate": 4.3743061062648696e-05,
+      "loss": 1.273,
+      "step": 23680
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.200197458267212,
+      "learning_rate": 4.374279672217817e-05,
+      "loss": 1.3884,
+      "step": 23681
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7927528619766235,
+      "learning_rate": 4.374253238170764e-05,
+      "loss": 1.2039,
+      "step": 23682
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.038358688354492,
+      "learning_rate": 4.374226804123712e-05,
+      "loss": 1.3854,
+      "step": 23683
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6773947477340698,
+      "learning_rate": 4.374200370076659e-05,
+      "loss": 1.2265,
+      "step": 23684
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9392170906066895,
+      "learning_rate": 4.3741739360296064e-05,
+      "loss": 1.1973,
+      "step": 23685
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0730743408203125,
+      "learning_rate": 4.374147501982554e-05,
+      "loss": 1.3896,
+      "step": 23686
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6553058624267578,
+      "learning_rate": 4.374121067935501e-05,
+      "loss": 2.1533,
+      "step": 23687
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.098327875137329,
+      "learning_rate": 4.3740946338884485e-05,
+      "loss": 1.158,
+      "step": 23688
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.259981393814087,
+      "learning_rate": 4.374068199841396e-05,
+      "loss": 1.2771,
+      "step": 23689
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.562703847885132,
+      "learning_rate": 4.374041765794343e-05,
+      "loss": 1.4161,
+      "step": 23690
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0346245765686035,
+      "learning_rate": 4.374015331747291e-05,
+      "loss": 1.3329,
+      "step": 23691
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.633968472480774,
+      "learning_rate": 4.373988897700238e-05,
+      "loss": 1.2854,
+      "step": 23692
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0054335594177246,
+      "learning_rate": 4.373962463653186e-05,
+      "loss": 1.263,
+      "step": 23693
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.2366981506347656,
+      "learning_rate": 4.3739360296061326e-05,
+      "loss": 1.5118,
+      "step": 23694
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0748934745788574,
+      "learning_rate": 4.3739095955590806e-05,
+      "loss": 1.2951,
+      "step": 23695
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9987982511520386,
+      "learning_rate": 4.373883161512028e-05,
+      "loss": 1.3041,
+      "step": 23696
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.361309051513672,
+      "learning_rate": 4.373856727464975e-05,
+      "loss": 2.4089,
+      "step": 23697
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0271263122558594,
+      "learning_rate": 4.373830293417923e-05,
+      "loss": 1.2266,
+      "step": 23698
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.784203052520752,
+      "learning_rate": 4.3738038593708694e-05,
+      "loss": 1.3902,
+      "step": 23699
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.226982355117798,
+      "learning_rate": 4.3737774253238174e-05,
+      "loss": 2.2285,
+      "step": 23700
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7342033386230469,
+      "learning_rate": 4.373750991276765e-05,
+      "loss": 1.0571,
+      "step": 23701
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9496631622314453,
+      "learning_rate": 4.373724557229712e-05,
+      "loss": 1.2717,
+      "step": 23702
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1062986850738525,
+      "learning_rate": 4.3736981231826595e-05,
+      "loss": 1.219,
+      "step": 23703
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.165821075439453,
+      "learning_rate": 4.373671689135607e-05,
+      "loss": 1.2679,
+      "step": 23704
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1307976245880127,
+      "learning_rate": 4.373645255088554e-05,
+      "loss": 1.2382,
+      "step": 23705
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9523345232009888,
+      "learning_rate": 4.3736188210415016e-05,
+      "loss": 1.2785,
+      "step": 23706
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7364517450332642,
+      "learning_rate": 4.373592386994449e-05,
+      "loss": 1.2201,
+      "step": 23707
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8194748163223267,
+      "learning_rate": 4.373565952947396e-05,
+      "loss": 1.2785,
+      "step": 23708
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8097065687179565,
+      "learning_rate": 4.3735395189003436e-05,
+      "loss": 1.214,
+      "step": 23709
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.752343654632568,
+      "learning_rate": 4.373513084853291e-05,
+      "loss": 1.342,
+      "step": 23710
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0342071056365967,
+      "learning_rate": 4.373486650806239e-05,
+      "loss": 1.2713,
+      "step": 23711
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.596792697906494,
+      "learning_rate": 4.373460216759186e-05,
+      "loss": 1.3525,
+      "step": 23712
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.4014577865600586,
+      "learning_rate": 4.373433782712134e-05,
+      "loss": 1.2541,
+      "step": 23713
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.308633327484131,
+      "learning_rate": 4.3734073486650804e-05,
+      "loss": 1.2294,
+      "step": 23714
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1313939094543457,
+      "learning_rate": 4.3733809146180285e-05,
+      "loss": 1.2136,
+      "step": 23715
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.066478967666626,
+      "learning_rate": 4.373354480570976e-05,
+      "loss": 1.3542,
+      "step": 23716
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6507419347763062,
+      "learning_rate": 4.373328046523923e-05,
+      "loss": 1.1965,
+      "step": 23717
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0137524604797363,
+      "learning_rate": 4.3733016124768705e-05,
+      "loss": 1.3293,
+      "step": 23718
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8443992137908936,
+      "learning_rate": 4.373275178429818e-05,
+      "loss": 1.2866,
+      "step": 23719
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.192091703414917,
+      "learning_rate": 4.373248744382765e-05,
+      "loss": 1.4125,
+      "step": 23720
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.7833473682403564,
+      "learning_rate": 4.3732223103357126e-05,
+      "loss": 1.4145,
+      "step": 23721
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.730189561843872,
+      "learning_rate": 4.37319587628866e-05,
+      "loss": 1.502,
+      "step": 23722
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.1352689266204834,
+      "learning_rate": 4.373169442241607e-05,
+      "loss": 1.3105,
+      "step": 23723
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9448808431625366,
+      "learning_rate": 4.373143008194555e-05,
+      "loss": 1.3837,
+      "step": 23724
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9045989513397217,
+      "learning_rate": 4.373116574147502e-05,
+      "loss": 1.2784,
+      "step": 23725
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.328625202178955,
+      "learning_rate": 4.37309014010045e-05,
+      "loss": 1.2858,
+      "step": 23726
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.3168108463287354,
+      "learning_rate": 4.373063706053397e-05,
+      "loss": 1.3102,
+      "step": 23727
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.763075828552246,
+      "learning_rate": 4.373037272006345e-05,
+      "loss": 1.4296,
+      "step": 23728
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.155259847640991,
+      "learning_rate": 4.3730108379592915e-05,
+      "loss": 1.3193,
+      "step": 23729
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.6808953285217285,
+      "learning_rate": 4.3729844039122395e-05,
+      "loss": 1.3949,
+      "step": 23730
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.782256841659546,
+      "learning_rate": 4.372957969865187e-05,
+      "loss": 1.3266,
+      "step": 23731
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.7969493865966797,
+      "learning_rate": 4.3729315358181335e-05,
+      "loss": 1.3324,
+      "step": 23732
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.723101019859314,
+      "learning_rate": 4.3729051017710816e-05,
+      "loss": 1.19,
+      "step": 23733
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9238134622573853,
+      "learning_rate": 4.372878667724028e-05,
+      "loss": 1.6907,
+      "step": 23734
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9323234558105469,
+      "learning_rate": 4.372852233676976e-05,
+      "loss": 1.2862,
+      "step": 23735
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8257719278335571,
+      "learning_rate": 4.3728257996299236e-05,
+      "loss": 1.2458,
+      "step": 23736
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.7965118885040283,
+      "learning_rate": 4.372799365582871e-05,
+      "loss": 1.5299,
+      "step": 23737
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7126305103302002,
+      "learning_rate": 4.3727729315358183e-05,
+      "loss": 1.233,
+      "step": 23738
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7950434684753418,
+      "learning_rate": 4.372746497488766e-05,
+      "loss": 1.1512,
+      "step": 23739
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9880940914154053,
+      "learning_rate": 4.372720063441713e-05,
+      "loss": 1.3616,
+      "step": 23740
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7878299951553345,
+      "learning_rate": 4.372693629394661e-05,
+      "loss": 1.2377,
+      "step": 23741
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3969626426696777,
+      "learning_rate": 4.372667195347608e-05,
+      "loss": 1.3104,
+      "step": 23742
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.1512293815612793,
+      "learning_rate": 4.372640761300555e-05,
+      "loss": 1.4527,
+      "step": 23743
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8925997018814087,
+      "learning_rate": 4.3726143272535025e-05,
+      "loss": 1.1637,
+      "step": 23744
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.229029893875122,
+      "learning_rate": 4.37258789320645e-05,
+      "loss": 1.3129,
+      "step": 23745
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.054957151412964,
+      "learning_rate": 4.372561459159398e-05,
+      "loss": 1.1596,
+      "step": 23746
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0891597270965576,
+      "learning_rate": 4.3725350251123446e-05,
+      "loss": 1.2948,
+      "step": 23747
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.001415967941284,
+      "learning_rate": 4.3725085910652926e-05,
+      "loss": 1.3694,
+      "step": 23748
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.012033224105835,
+      "learning_rate": 4.372482157018239e-05,
+      "loss": 1.416,
+      "step": 23749
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9632267951965332,
+      "learning_rate": 4.372455722971187e-05,
+      "loss": 1.988,
+      "step": 23750
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8252694606781006,
+      "learning_rate": 4.372429288924135e-05,
+      "loss": 1.2861,
+      "step": 23751
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.3126847743988037,
+      "learning_rate": 4.372402854877082e-05,
+      "loss": 1.6693,
+      "step": 23752
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8861668109893799,
+      "learning_rate": 4.3723764208300294e-05,
+      "loss": 1.3895,
+      "step": 23753
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.133301258087158,
+      "learning_rate": 4.372349986782977e-05,
+      "loss": 1.2839,
+      "step": 23754
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7150099277496338,
+      "learning_rate": 4.372323552735924e-05,
+      "loss": 1.2916,
+      "step": 23755
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.893803119659424,
+      "learning_rate": 4.3722971186888715e-05,
+      "loss": 1.3565,
+      "step": 23756
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7134356498718262,
+      "learning_rate": 4.372270684641819e-05,
+      "loss": 1.3255,
+      "step": 23757
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.34303879737854,
+      "learning_rate": 4.372244250594766e-05,
+      "loss": 1.238,
+      "step": 23758
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9984095096588135,
+      "learning_rate": 4.3722178165477135e-05,
+      "loss": 1.3171,
+      "step": 23759
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7337887287139893,
+      "learning_rate": 4.372191382500661e-05,
+      "loss": 1.1012,
+      "step": 23760
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.003955125808716,
+      "learning_rate": 4.372164948453609e-05,
+      "loss": 1.2276,
+      "step": 23761
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.969435214996338,
+      "learning_rate": 4.3721385144065556e-05,
+      "loss": 1.5655,
+      "step": 23762
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.107691764831543,
+      "learning_rate": 4.3721120803595036e-05,
+      "loss": 1.3017,
+      "step": 23763
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.025423288345337,
+      "learning_rate": 4.37208564631245e-05,
+      "loss": 1.2738,
+      "step": 23764
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0190181732177734,
+      "learning_rate": 4.3720592122653983e-05,
+      "loss": 1.4715,
+      "step": 23765
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.8667454719543457,
+      "learning_rate": 4.372032778218346e-05,
+      "loss": 1.3052,
+      "step": 23766
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9258716106414795,
+      "learning_rate": 4.372006344171293e-05,
+      "loss": 1.1862,
+      "step": 23767
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.5796809196472168,
+      "learning_rate": 4.3719799101242404e-05,
+      "loss": 1.208,
+      "step": 23768
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.2610039710998535,
+      "learning_rate": 4.371953476077187e-05,
+      "loss": 1.391,
+      "step": 23769
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.917699098587036,
+      "learning_rate": 4.371927042030135e-05,
+      "loss": 1.4228,
+      "step": 23770
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.149972677230835,
+      "learning_rate": 4.3719006079830825e-05,
+      "loss": 1.3345,
+      "step": 23771
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8120543956756592,
+      "learning_rate": 4.37187417393603e-05,
+      "loss": 1.3702,
+      "step": 23772
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9435616731643677,
+      "learning_rate": 4.371847739888977e-05,
+      "loss": 1.3075,
+      "step": 23773
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.1834793090820312,
+      "learning_rate": 4.3718213058419246e-05,
+      "loss": 1.2398,
+      "step": 23774
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.881582260131836,
+      "learning_rate": 4.371794871794872e-05,
+      "loss": 1.2254,
+      "step": 23775
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.939111590385437,
+      "learning_rate": 4.37176843774782e-05,
+      "loss": 1.2804,
+      "step": 23776
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.6913135051727295,
+      "learning_rate": 4.3717420037007666e-05,
+      "loss": 1.2064,
+      "step": 23777
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.999497413635254,
+      "learning_rate": 4.371715569653714e-05,
+      "loss": 1.3018,
+      "step": 23778
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.127453565597534,
+      "learning_rate": 4.3716891356066613e-05,
+      "loss": 1.3233,
+      "step": 23779
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8859254121780396,
+      "learning_rate": 4.371662701559609e-05,
+      "loss": 1.3504,
+      "step": 23780
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.2662250995635986,
+      "learning_rate": 4.371636267512557e-05,
+      "loss": 1.3816,
+      "step": 23781
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.371325969696045,
+      "learning_rate": 4.3716098334655034e-05,
+      "loss": 1.4736,
+      "step": 23782
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0685343742370605,
+      "learning_rate": 4.3715833994184515e-05,
+      "loss": 1.4739,
+      "step": 23783
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.470609426498413,
+      "learning_rate": 4.371556965371398e-05,
+      "loss": 1.207,
+      "step": 23784
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.1111321449279785,
+      "learning_rate": 4.371530531324346e-05,
+      "loss": 1.3824,
+      "step": 23785
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.009408950805664,
+      "learning_rate": 4.3715040972772935e-05,
+      "loss": 1.253,
+      "step": 23786
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3625385761260986,
+      "learning_rate": 4.371477663230241e-05,
+      "loss": 1.4505,
+      "step": 23787
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.66105318069458,
+      "learning_rate": 4.371451229183188e-05,
+      "loss": 1.2924,
+      "step": 23788
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9116755723953247,
+      "learning_rate": 4.3714247951361356e-05,
+      "loss": 1.2118,
+      "step": 23789
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.066074848175049,
+      "learning_rate": 4.371398361089083e-05,
+      "loss": 1.3995,
+      "step": 23790
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.4139063358306885,
+      "learning_rate": 4.37137192704203e-05,
+      "loss": 1.5231,
+      "step": 23791
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.8635013103485107,
+      "learning_rate": 4.371345492994978e-05,
+      "loss": 1.3924,
+      "step": 23792
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.1040287017822266,
+      "learning_rate": 4.371319058947925e-05,
+      "loss": 1.3452,
+      "step": 23793
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.010478973388672,
+      "learning_rate": 4.3712926249008724e-05,
+      "loss": 1.3081,
+      "step": 23794
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.66539466381073,
+      "learning_rate": 4.37126619085382e-05,
+      "loss": 1.2505,
+      "step": 23795
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.132021903991699,
+      "learning_rate": 4.371239756806768e-05,
+      "loss": 1.5809,
+      "step": 23796
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.076566696166992,
+      "learning_rate": 4.3712133227597144e-05,
+      "loss": 1.3313,
+      "step": 23797
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.957641363143921,
+      "learning_rate": 4.3711868887126625e-05,
+      "loss": 1.3259,
+      "step": 23798
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0202975273132324,
+      "learning_rate": 4.371160454665609e-05,
+      "loss": 1.2063,
+      "step": 23799
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.145059108734131,
+      "learning_rate": 4.371134020618557e-05,
+      "loss": 1.2117,
+      "step": 23800
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0549733638763428,
+      "learning_rate": 4.3711075865715046e-05,
+      "loss": 1.2503,
+      "step": 23801
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.185748338699341,
+      "learning_rate": 4.371081152524452e-05,
+      "loss": 1.3118,
+      "step": 23802
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.134967565536499,
+      "learning_rate": 4.371054718477399e-05,
+      "loss": 1.2714,
+      "step": 23803
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0073461532592773,
+      "learning_rate": 4.371028284430346e-05,
+      "loss": 1.6663,
+      "step": 23804
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0749340057373047,
+      "learning_rate": 4.371001850383294e-05,
+      "loss": 1.4138,
+      "step": 23805
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.141444444656372,
+      "learning_rate": 4.3709754163362413e-05,
+      "loss": 1.4351,
+      "step": 23806
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.776088237762451,
+      "learning_rate": 4.370948982289189e-05,
+      "loss": 1.3504,
+      "step": 23807
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6268733739852905,
+      "learning_rate": 4.370922548242136e-05,
+      "loss": 1.19,
+      "step": 23808
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8584635257720947,
+      "learning_rate": 4.3708961141950834e-05,
+      "loss": 1.3005,
+      "step": 23809
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.9781675338745117,
+      "learning_rate": 4.370869680148031e-05,
+      "loss": 1.5116,
+      "step": 23810
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.212329387664795,
+      "learning_rate": 4.370843246100979e-05,
+      "loss": 1.3173,
+      "step": 23811
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.341794729232788,
+      "learning_rate": 4.3708168120539255e-05,
+      "loss": 1.4097,
+      "step": 23812
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.529325008392334,
+      "learning_rate": 4.3707903780068735e-05,
+      "loss": 1.4456,
+      "step": 23813
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.899115800857544,
+      "learning_rate": 4.37076394395982e-05,
+      "loss": 1.165,
+      "step": 23814
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.047917127609253,
+      "learning_rate": 4.3707375099127676e-05,
+      "loss": 1.3269,
+      "step": 23815
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3611819744110107,
+      "learning_rate": 4.3707110758657156e-05,
+      "loss": 1.7305,
+      "step": 23816
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8737068176269531,
+      "learning_rate": 4.370684641818662e-05,
+      "loss": 1.2075,
+      "step": 23817
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0729053020477295,
+      "learning_rate": 4.37065820777161e-05,
+      "loss": 1.279,
+      "step": 23818
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.4197051525115967,
+      "learning_rate": 4.370631773724557e-05,
+      "loss": 1.9019,
+      "step": 23819
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.178624391555786,
+      "learning_rate": 4.370605339677505e-05,
+      "loss": 1.3176,
+      "step": 23820
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.894615650177002,
+      "learning_rate": 4.3705789056304524e-05,
+      "loss": 1.2562,
+      "step": 23821
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9897123575210571,
+      "learning_rate": 4.3705524715834e-05,
+      "loss": 1.3451,
+      "step": 23822
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7660728693008423,
+      "learning_rate": 4.370526037536347e-05,
+      "loss": 1.2339,
+      "step": 23823
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.2298777103424072,
+      "learning_rate": 4.3704996034892944e-05,
+      "loss": 1.39,
+      "step": 23824
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.395660400390625,
+      "learning_rate": 4.370473169442242e-05,
+      "loss": 1.3661,
+      "step": 23825
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.043123722076416,
+      "learning_rate": 4.370446735395189e-05,
+      "loss": 1.167,
+      "step": 23826
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.775017023086548,
+      "learning_rate": 4.3704203013481365e-05,
+      "loss": 1.4245,
+      "step": 23827
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8786271810531616,
+      "learning_rate": 4.370393867301084e-05,
+      "loss": 2.0358,
+      "step": 23828
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.260178565979004,
+      "learning_rate": 4.370367433254031e-05,
+      "loss": 1.3593,
+      "step": 23829
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.73977530002594,
+      "learning_rate": 4.3703409992069786e-05,
+      "loss": 1.1575,
+      "step": 23830
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.17919659614563,
+      "learning_rate": 4.3703145651599266e-05,
+      "loss": 1.4086,
+      "step": 23831
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.003845691680908,
+      "learning_rate": 4.370288131112873e-05,
+      "loss": 1.3426,
+      "step": 23832
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.0318899154663086,
+      "learning_rate": 4.370261697065821e-05,
+      "loss": 1.2732,
+      "step": 23833
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9431960582733154,
+      "learning_rate": 4.370235263018768e-05,
+      "loss": 1.309,
+      "step": 23834
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.036855697631836,
+      "learning_rate": 4.370208828971716e-05,
+      "loss": 1.236,
+      "step": 23835
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.424212694168091,
+      "learning_rate": 4.3701823949246634e-05,
+      "loss": 1.3656,
+      "step": 23836
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.602353096008301,
+      "learning_rate": 4.370155960877611e-05,
+      "loss": 1.2421,
+      "step": 23837
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.1604673862457275,
+      "learning_rate": 4.370129526830558e-05,
+      "loss": 1.4723,
+      "step": 23838
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.332308292388916,
+      "learning_rate": 4.3701030927835055e-05,
+      "loss": 1.3839,
+      "step": 23839
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.958528995513916,
+      "learning_rate": 4.370076658736453e-05,
+      "loss": 1.3058,
+      "step": 23840
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9772638082504272,
+      "learning_rate": 4.3700502246894e-05,
+      "loss": 1.3816,
+      "step": 23841
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8830292224884033,
+      "learning_rate": 4.3700237906423476e-05,
+      "loss": 1.2947,
+      "step": 23842
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0467257499694824,
+      "learning_rate": 4.369997356595295e-05,
+      "loss": 1.2158,
+      "step": 23843
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.940567970275879,
+      "learning_rate": 4.369970922548242e-05,
+      "loss": 1.2726,
+      "step": 23844
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8171842098236084,
+      "learning_rate": 4.3699444885011896e-05,
+      "loss": 1.2703,
+      "step": 23845
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.2666730880737305,
+      "learning_rate": 4.3699180544541377e-05,
+      "loss": 1.4276,
+      "step": 23846
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9337048530578613,
+      "learning_rate": 4.369891620407084e-05,
+      "loss": 1.2234,
+      "step": 23847
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9041974544525146,
+      "learning_rate": 4.3698651863600324e-05,
+      "loss": 1.3753,
+      "step": 23848
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6749976873397827,
+      "learning_rate": 4.369838752312979e-05,
+      "loss": 1.2638,
+      "step": 23849
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9174377918243408,
+      "learning_rate": 4.3698123182659264e-05,
+      "loss": 1.2774,
+      "step": 23850
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.2036495208740234,
+      "learning_rate": 4.3697858842188744e-05,
+      "loss": 1.3028,
+      "step": 23851
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.185506582260132,
+      "learning_rate": 4.369759450171821e-05,
+      "loss": 1.3928,
+      "step": 23852
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.0834743976593018,
+      "learning_rate": 4.369733016124769e-05,
+      "loss": 1.258,
+      "step": 23853
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3633692264556885,
+      "learning_rate": 4.369706582077716e-05,
+      "loss": 1.4072,
+      "step": 23854
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0453624725341797,
+      "learning_rate": 4.369680148030664e-05,
+      "loss": 1.3442,
+      "step": 23855
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.155362606048584,
+      "learning_rate": 4.369653713983611e-05,
+      "loss": 1.2661,
+      "step": 23856
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.729585647583008,
+      "learning_rate": 4.3696272799365586e-05,
+      "loss": 1.3831,
+      "step": 23857
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.777511477470398,
+      "learning_rate": 4.369600845889506e-05,
+      "loss": 1.2819,
+      "step": 23858
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.1353349685668945,
+      "learning_rate": 4.369574411842453e-05,
+      "loss": 1.316,
+      "step": 23859
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.059832811355591,
+      "learning_rate": 4.3695479777954007e-05,
+      "loss": 1.3337,
+      "step": 23860
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.321807622909546,
+      "learning_rate": 4.369521543748348e-05,
+      "loss": 1.3787,
+      "step": 23861
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.196791410446167,
+      "learning_rate": 4.3694951097012954e-05,
+      "loss": 1.3341,
+      "step": 23862
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3158881664276123,
+      "learning_rate": 4.369468675654243e-05,
+      "loss": 1.308,
+      "step": 23863
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.884840488433838,
+      "learning_rate": 4.36944224160719e-05,
+      "loss": 1.1881,
+      "step": 23864
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.107417583465576,
+      "learning_rate": 4.3694158075601374e-05,
+      "loss": 1.3299,
+      "step": 23865
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7015118598937988,
+      "learning_rate": 4.3693893735130855e-05,
+      "loss": 1.1974,
+      "step": 23866
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.359849691390991,
+      "learning_rate": 4.369362939466032e-05,
+      "loss": 1.4988,
+      "step": 23867
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.5554215908050537,
+      "learning_rate": 4.36933650541898e-05,
+      "loss": 1.311,
+      "step": 23868
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.6855616569519043,
+      "learning_rate": 4.369310071371927e-05,
+      "loss": 1.2663,
+      "step": 23869
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8519408702850342,
+      "learning_rate": 4.369283637324875e-05,
+      "loss": 1.1561,
+      "step": 23870
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.1910645961761475,
+      "learning_rate": 4.369257203277822e-05,
+      "loss": 1.2699,
+      "step": 23871
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.6295881271362305,
+      "learning_rate": 4.3692307692307696e-05,
+      "loss": 1.3544,
+      "step": 23872
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.2944273948669434,
+      "learning_rate": 4.369204335183717e-05,
+      "loss": 1.3494,
+      "step": 23873
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0226023197174072,
+      "learning_rate": 4.369177901136664e-05,
+      "loss": 1.235,
+      "step": 23874
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.339977502822876,
+      "learning_rate": 4.369151467089612e-05,
+      "loss": 1.4222,
+      "step": 23875
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6292914152145386,
+      "learning_rate": 4.369125033042559e-05,
+      "loss": 1.1903,
+      "step": 23876
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.4450504779815674,
+      "learning_rate": 4.3690985989955064e-05,
+      "loss": 1.4266,
+      "step": 23877
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9580496549606323,
+      "learning_rate": 4.369072164948454e-05,
+      "loss": 1.2649,
+      "step": 23878
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.212315797805786,
+      "learning_rate": 4.369045730901401e-05,
+      "loss": 1.4546,
+      "step": 23879
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6807821989059448,
+      "learning_rate": 4.3690192968543485e-05,
+      "loss": 1.1817,
+      "step": 23880
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8767834901809692,
+      "learning_rate": 4.3689928628072965e-05,
+      "loss": 1.0995,
+      "step": 23881
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.216259002685547,
+      "learning_rate": 4.368966428760243e-05,
+      "loss": 1.3285,
+      "step": 23882
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3924098014831543,
+      "learning_rate": 4.368939994713191e-05,
+      "loss": 1.38,
+      "step": 23883
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.2264440059661865,
+      "learning_rate": 4.368913560666138e-05,
+      "loss": 1.2839,
+      "step": 23884
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7992924451828003,
+      "learning_rate": 4.368887126619086e-05,
+      "loss": 1.1273,
+      "step": 23885
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.832786202430725,
+      "learning_rate": 4.368860692572033e-05,
+      "loss": 1.3091,
+      "step": 23886
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.6117889881134033,
+      "learning_rate": 4.36883425852498e-05,
+      "loss": 1.424,
+      "step": 23887
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0866141319274902,
+      "learning_rate": 4.368807824477928e-05,
+      "loss": 1.3406,
+      "step": 23888
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0010316371917725,
+      "learning_rate": 4.368781390430875e-05,
+      "loss": 1.2642,
+      "step": 23889
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8452847003936768,
+      "learning_rate": 4.368754956383823e-05,
+      "loss": 1.2989,
+      "step": 23890
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3648183345794678,
+      "learning_rate": 4.36872852233677e-05,
+      "loss": 1.3844,
+      "step": 23891
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.6864264011383057,
+      "learning_rate": 4.3687020882897174e-05,
+      "loss": 1.4177,
+      "step": 23892
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.7515580654144287,
+      "learning_rate": 4.368675654242665e-05,
+      "loss": 1.4326,
+      "step": 23893
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.2035861015319824,
+      "learning_rate": 4.368649220195612e-05,
+      "loss": 2.0773,
+      "step": 23894
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.4929325580596924,
+      "learning_rate": 4.3686227861485595e-05,
+      "loss": 1.3784,
+      "step": 23895
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.120407819747925,
+      "learning_rate": 4.368596352101507e-05,
+      "loss": 1.3252,
+      "step": 23896
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.240412950515747,
+      "learning_rate": 4.368569918054454e-05,
+      "loss": 1.2831,
+      "step": 23897
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.4036033153533936,
+      "learning_rate": 4.3685434840074016e-05,
+      "loss": 1.337,
+      "step": 23898
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6564817428588867,
+      "learning_rate": 4.368517049960349e-05,
+      "loss": 1.2333,
+      "step": 23899
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.895548701286316,
+      "learning_rate": 4.368490615913296e-05,
+      "loss": 1.3631,
+      "step": 23900
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8035662174224854,
+      "learning_rate": 4.368464181866244e-05,
+      "loss": 1.2735,
+      "step": 23901
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6871789693832397,
+      "learning_rate": 4.368437747819191e-05,
+      "loss": 1.1577,
+      "step": 23902
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.4222917556762695,
+      "learning_rate": 4.368411313772139e-05,
+      "loss": 1.2659,
+      "step": 23903
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.037891149520874,
+      "learning_rate": 4.368384879725086e-05,
+      "loss": 1.2745,
+      "step": 23904
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.149791717529297,
+      "learning_rate": 4.368358445678034e-05,
+      "loss": 1.267,
+      "step": 23905
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9297410249710083,
+      "learning_rate": 4.368332011630981e-05,
+      "loss": 1.275,
+      "step": 23906
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.5855575799942017,
+      "learning_rate": 4.3683055775839285e-05,
+      "loss": 1.2627,
+      "step": 23907
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0971970558166504,
+      "learning_rate": 4.368279143536876e-05,
+      "loss": 1.318,
+      "step": 23908
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9443100690841675,
+      "learning_rate": 4.368252709489823e-05,
+      "loss": 1.3091,
+      "step": 23909
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8493216037750244,
+      "learning_rate": 4.3682262754427705e-05,
+      "loss": 1.2262,
+      "step": 23910
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.280888080596924,
+      "learning_rate": 4.368199841395718e-05,
+      "loss": 1.3675,
+      "step": 23911
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8341377973556519,
+      "learning_rate": 4.368173407348665e-05,
+      "loss": 1.2611,
+      "step": 23912
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.242636203765869,
+      "learning_rate": 4.3681469733016126e-05,
+      "loss": 1.4416,
+      "step": 23913
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.544985294342041,
+      "learning_rate": 4.36812053925456e-05,
+      "loss": 1.4041,
+      "step": 23914
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.360694408416748,
+      "learning_rate": 4.368094105207507e-05,
+      "loss": 1.4964,
+      "step": 23915
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9992643594741821,
+      "learning_rate": 4.3680676711604554e-05,
+      "loss": 1.4168,
+      "step": 23916
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9169059991836548,
+      "learning_rate": 4.368041237113402e-05,
+      "loss": 1.3167,
+      "step": 23917
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.826429843902588,
+      "learning_rate": 4.36801480306635e-05,
+      "loss": 1.3442,
+      "step": 23918
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.4255049228668213,
+      "learning_rate": 4.367988369019297e-05,
+      "loss": 1.2479,
+      "step": 23919
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9413329362869263,
+      "learning_rate": 4.367961934972245e-05,
+      "loss": 1.2906,
+      "step": 23920
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7827595472335815,
+      "learning_rate": 4.367935500925192e-05,
+      "loss": 1.3018,
+      "step": 23921
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.929180383682251,
+      "learning_rate": 4.367909066878139e-05,
+      "loss": 1.3591,
+      "step": 23922
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3199872970581055,
+      "learning_rate": 4.367882632831087e-05,
+      "loss": 1.5803,
+      "step": 23923
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7925280332565308,
+      "learning_rate": 4.3678561987840335e-05,
+      "loss": 1.4414,
+      "step": 23924
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.040761947631836,
+      "learning_rate": 4.3678297647369816e-05,
+      "loss": 1.2189,
+      "step": 23925
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8137810230255127,
+      "learning_rate": 4.367803330689929e-05,
+      "loss": 1.5111,
+      "step": 23926
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.975938081741333,
+      "learning_rate": 4.367776896642876e-05,
+      "loss": 1.3055,
+      "step": 23927
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9917175769805908,
+      "learning_rate": 4.3677504625958236e-05,
+      "loss": 1.2511,
+      "step": 23928
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.101008892059326,
+      "learning_rate": 4.367724028548771e-05,
+      "loss": 1.2032,
+      "step": 23929
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.05112361907959,
+      "learning_rate": 4.3676975945017184e-05,
+      "loss": 1.2149,
+      "step": 23930
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9354145526885986,
+      "learning_rate": 4.3676711604546664e-05,
+      "loss": 1.2783,
+      "step": 23931
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4073240756988525,
+      "learning_rate": 4.367644726407613e-05,
+      "loss": 1.2439,
+      "step": 23932
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.310398817062378,
+      "learning_rate": 4.3676182923605604e-05,
+      "loss": 1.3545,
+      "step": 23933
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8008701801300049,
+      "learning_rate": 4.367591858313508e-05,
+      "loss": 1.2164,
+      "step": 23934
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.262599229812622,
+      "learning_rate": 4.367565424266455e-05,
+      "loss": 1.2831,
+      "step": 23935
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.541510820388794,
+      "learning_rate": 4.367538990219403e-05,
+      "loss": 1.1428,
+      "step": 23936
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7958416938781738,
+      "learning_rate": 4.36751255617235e-05,
+      "loss": 1.2428,
+      "step": 23937
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.155733346939087,
+      "learning_rate": 4.367486122125298e-05,
+      "loss": 1.2489,
+      "step": 23938
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.157243013381958,
+      "learning_rate": 4.3674596880782446e-05,
+      "loss": 1.373,
+      "step": 23939
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.83722722530365,
+      "learning_rate": 4.3674332540311926e-05,
+      "loss": 1.1846,
+      "step": 23940
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1220247745513916,
+      "learning_rate": 4.36740681998414e-05,
+      "loss": 1.4588,
+      "step": 23941
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.422042965888977,
+      "learning_rate": 4.367380385937087e-05,
+      "loss": 1.2394,
+      "step": 23942
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.290408134460449,
+      "learning_rate": 4.367353951890035e-05,
+      "loss": 1.4473,
+      "step": 23943
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.4988411664962769,
+      "learning_rate": 4.367327517842982e-05,
+      "loss": 1.4725,
+      "step": 23944
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9696377515792847,
+      "learning_rate": 4.3673010837959294e-05,
+      "loss": 1.3146,
+      "step": 23945
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.269540309906006,
+      "learning_rate": 4.367274649748877e-05,
+      "loss": 1.2991,
+      "step": 23946
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1728198528289795,
+      "learning_rate": 4.367248215701824e-05,
+      "loss": 1.4554,
+      "step": 23947
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4755642414093018,
+      "learning_rate": 4.3672217816547715e-05,
+      "loss": 1.2855,
+      "step": 23948
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7971066236495972,
+      "learning_rate": 4.367195347607719e-05,
+      "loss": 1.3768,
+      "step": 23949
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.149319887161255,
+      "learning_rate": 4.367168913560666e-05,
+      "loss": 1.3779,
+      "step": 23950
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8278025388717651,
+      "learning_rate": 4.367142479513614e-05,
+      "loss": 1.2761,
+      "step": 23951
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.7054851055145264,
+      "learning_rate": 4.367116045466561e-05,
+      "loss": 1.5729,
+      "step": 23952
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.897642135620117,
+      "learning_rate": 4.367089611419509e-05,
+      "loss": 1.3207,
+      "step": 23953
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.9488558769226074,
+      "learning_rate": 4.3670631773724556e-05,
+      "loss": 1.3602,
+      "step": 23954
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4458823204040527,
+      "learning_rate": 4.3670367433254036e-05,
+      "loss": 1.3637,
+      "step": 23955
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.23577880859375,
+      "learning_rate": 4.367010309278351e-05,
+      "loss": 1.3082,
+      "step": 23956
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0351991653442383,
+      "learning_rate": 4.3669838752312984e-05,
+      "loss": 1.1659,
+      "step": 23957
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.3124265670776367,
+      "learning_rate": 4.366957441184246e-05,
+      "loss": 1.3925,
+      "step": 23958
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9896782636642456,
+      "learning_rate": 4.3669310071371924e-05,
+      "loss": 1.2856,
+      "step": 23959
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9532026052474976,
+      "learning_rate": 4.3669045730901404e-05,
+      "loss": 1.336,
+      "step": 23960
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5898381471633911,
+      "learning_rate": 4.366878139043088e-05,
+      "loss": 1.2937,
+      "step": 23961
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.869276523590088,
+      "learning_rate": 4.366851704996035e-05,
+      "loss": 2.2045,
+      "step": 23962
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1234710216522217,
+      "learning_rate": 4.3668252709489825e-05,
+      "loss": 1.3186,
+      "step": 23963
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7254501581192017,
+      "learning_rate": 4.36679883690193e-05,
+      "loss": 1.1298,
+      "step": 23964
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.218310594558716,
+      "learning_rate": 4.366772402854877e-05,
+      "loss": 1.2275,
+      "step": 23965
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8115571737289429,
+      "learning_rate": 4.366745968807825e-05,
+      "loss": 1.2565,
+      "step": 23966
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.014965534210205,
+      "learning_rate": 4.366719534760772e-05,
+      "loss": 1.3612,
+      "step": 23967
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.977774977684021,
+      "learning_rate": 4.366693100713719e-05,
+      "loss": 1.2886,
+      "step": 23968
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.038195848464966,
+      "learning_rate": 4.3666666666666666e-05,
+      "loss": 1.4571,
+      "step": 23969
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9014649391174316,
+      "learning_rate": 4.366640232619614e-05,
+      "loss": 1.2267,
+      "step": 23970
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.2229437828063965,
+      "learning_rate": 4.366613798572562e-05,
+      "loss": 1.4082,
+      "step": 23971
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8132011890411377,
+      "learning_rate": 4.366587364525509e-05,
+      "loss": 1.2879,
+      "step": 23972
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7421177625656128,
+      "learning_rate": 4.366560930478457e-05,
+      "loss": 1.1174,
+      "step": 23973
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.2529993057250977,
+      "learning_rate": 4.3665344964314034e-05,
+      "loss": 1.3617,
+      "step": 23974
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.040891408920288,
+      "learning_rate": 4.3665080623843515e-05,
+      "loss": 1.4491,
+      "step": 23975
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.08107590675354,
+      "learning_rate": 4.366481628337299e-05,
+      "loss": 1.3283,
+      "step": 23976
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5764038562774658,
+      "learning_rate": 4.366455194290246e-05,
+      "loss": 1.2675,
+      "step": 23977
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.015109062194824,
+      "learning_rate": 4.3664287602431935e-05,
+      "loss": 1.1458,
+      "step": 23978
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.277906656265259,
+      "learning_rate": 4.366402326196141e-05,
+      "loss": 1.4199,
+      "step": 23979
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1404566764831543,
+      "learning_rate": 4.366375892149088e-05,
+      "loss": 1.3473,
+      "step": 23980
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7260044813156128,
+      "learning_rate": 4.3663494581020356e-05,
+      "loss": 1.2375,
+      "step": 23981
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7724366188049316,
+      "learning_rate": 4.366323024054983e-05,
+      "loss": 1.1521,
+      "step": 23982
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4581003189086914,
+      "learning_rate": 4.36629659000793e-05,
+      "loss": 1.3273,
+      "step": 23983
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.582932949066162,
+      "learning_rate": 4.366270155960878e-05,
+      "loss": 1.364,
+      "step": 23984
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.138357400894165,
+      "learning_rate": 4.366243721913825e-05,
+      "loss": 1.4345,
+      "step": 23985
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.1378333568573,
+      "learning_rate": 4.366217287866773e-05,
+      "loss": 1.408,
+      "step": 23986
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.968565583229065,
+      "learning_rate": 4.36619085381972e-05,
+      "loss": 1.4283,
+      "step": 23987
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.3554649353027344,
+      "learning_rate": 4.366164419772668e-05,
+      "loss": 2.1303,
+      "step": 23988
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.521568536758423,
+      "learning_rate": 4.3661379857256145e-05,
+      "loss": 1.514,
+      "step": 23989
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.775134563446045,
+      "learning_rate": 4.3661115516785625e-05,
+      "loss": 1.216,
+      "step": 23990
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.6432826519012451,
+      "learning_rate": 4.36608511763151e-05,
+      "loss": 1.776,
+      "step": 23991
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9606361389160156,
+      "learning_rate": 4.366058683584457e-05,
+      "loss": 1.375,
+      "step": 23992
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.7040815353393555,
+      "learning_rate": 4.3660322495374046e-05,
+      "loss": 1.3588,
+      "step": 23993
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1917173862457275,
+      "learning_rate": 4.366005815490352e-05,
+      "loss": 1.3864,
+      "step": 23994
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.198885440826416,
+      "learning_rate": 4.365979381443299e-05,
+      "loss": 1.2628,
+      "step": 23995
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.797975778579712,
+      "learning_rate": 4.3659529473962466e-05,
+      "loss": 1.5033,
+      "step": 23996
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0596022605895996,
+      "learning_rate": 4.365926513349194e-05,
+      "loss": 1.2687,
+      "step": 23997
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.004011631011963,
+      "learning_rate": 4.3659000793021414e-05,
+      "loss": 1.1454,
+      "step": 23998
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8596789836883545,
+      "learning_rate": 4.365873645255089e-05,
+      "loss": 1.4332,
+      "step": 23999
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.053532838821411,
+      "learning_rate": 4.365847211208036e-05,
+      "loss": 1.4311,
+      "step": 24000
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.8001415729522705,
+      "learning_rate": 4.365820777160984e-05,
+      "loss": 1.3087,
+      "step": 24001
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7129772901535034,
+      "learning_rate": 4.365794343113931e-05,
+      "loss": 1.1983,
+      "step": 24002
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.7189295291900635,
+      "learning_rate": 4.365767909066879e-05,
+      "loss": 1.3512,
+      "step": 24003
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.420809030532837,
+      "learning_rate": 4.3657414750198255e-05,
+      "loss": 1.3667,
+      "step": 24004
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8803297281265259,
+      "learning_rate": 4.365715040972773e-05,
+      "loss": 1.2684,
+      "step": 24005
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7513821125030518,
+      "learning_rate": 4.365688606925721e-05,
+      "loss": 1.9512,
+      "step": 24006
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.6246254444122314,
+      "learning_rate": 4.3656621728786676e-05,
+      "loss": 1.2881,
+      "step": 24007
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7806583642959595,
+      "learning_rate": 4.3656357388316156e-05,
+      "loss": 1.3064,
+      "step": 24008
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0695316791534424,
+      "learning_rate": 4.365609304784562e-05,
+      "loss": 1.2628,
+      "step": 24009
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.189655065536499,
+      "learning_rate": 4.36558287073751e-05,
+      "loss": 1.2234,
+      "step": 24010
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.323418378829956,
+      "learning_rate": 4.365556436690458e-05,
+      "loss": 1.3415,
+      "step": 24011
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4658138751983643,
+      "learning_rate": 4.365530002643405e-05,
+      "loss": 1.3944,
+      "step": 24012
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9525350332260132,
+      "learning_rate": 4.3655035685963524e-05,
+      "loss": 1.3298,
+      "step": 24013
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.6712546348571777,
+      "learning_rate": 4.3654771345493e-05,
+      "loss": 1.2633,
+      "step": 24014
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.052133560180664,
+      "learning_rate": 4.365450700502247e-05,
+      "loss": 1.2724,
+      "step": 24015
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7596561908721924,
+      "learning_rate": 4.3654242664551945e-05,
+      "loss": 1.2794,
+      "step": 24016
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8298367261886597,
+      "learning_rate": 4.365397832408142e-05,
+      "loss": 1.2829,
+      "step": 24017
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7817610502243042,
+      "learning_rate": 4.365371398361089e-05,
+      "loss": 1.1906,
+      "step": 24018
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.601582646369934,
+      "learning_rate": 4.3653449643140365e-05,
+      "loss": 3.0186,
+      "step": 24019
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7633382081985474,
+      "learning_rate": 4.365318530266984e-05,
+      "loss": 1.2082,
+      "step": 24020
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8347182273864746,
+      "learning_rate": 4.365292096219932e-05,
+      "loss": 1.1928,
+      "step": 24021
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8048430681228638,
+      "learning_rate": 4.3652656621728786e-05,
+      "loss": 1.1315,
+      "step": 24022
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.6793702840805054,
+      "learning_rate": 4.3652392281258266e-05,
+      "loss": 1.2312,
+      "step": 24023
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.977169156074524,
+      "learning_rate": 4.365212794078773e-05,
+      "loss": 1.2879,
+      "step": 24024
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4598913192749023,
+      "learning_rate": 4.3651863600317214e-05,
+      "loss": 1.4679,
+      "step": 24025
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8339194059371948,
+      "learning_rate": 4.365159925984669e-05,
+      "loss": 1.294,
+      "step": 24026
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.5794517993927,
+      "learning_rate": 4.365133491937616e-05,
+      "loss": 1.324,
+      "step": 24027
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0334861278533936,
+      "learning_rate": 4.3651070578905634e-05,
+      "loss": 1.1865,
+      "step": 24028
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5220441818237305,
+      "learning_rate": 4.365080623843511e-05,
+      "loss": 2.7546,
+      "step": 24029
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.5740766525268555,
+      "learning_rate": 4.365054189796458e-05,
+      "loss": 1.4417,
+      "step": 24030
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.2562928199768066,
+      "learning_rate": 4.3650277557494055e-05,
+      "loss": 1.337,
+      "step": 24031
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9729108810424805,
+      "learning_rate": 4.365001321702353e-05,
+      "loss": 1.2981,
+      "step": 24032
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9063010215759277,
+      "learning_rate": 4.3649748876553e-05,
+      "loss": 1.3781,
+      "step": 24033
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5038471221923828,
+      "learning_rate": 4.3649484536082476e-05,
+      "loss": 1.1743,
+      "step": 24034
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7748804092407227,
+      "learning_rate": 4.364922019561195e-05,
+      "loss": 1.2739,
+      "step": 24035
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9230506420135498,
+      "learning_rate": 4.364895585514143e-05,
+      "loss": 1.1688,
+      "step": 24036
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9294841289520264,
+      "learning_rate": 4.3648691514670896e-05,
+      "loss": 1.134,
+      "step": 24037
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8736169338226318,
+      "learning_rate": 4.364842717420038e-05,
+      "loss": 1.266,
+      "step": 24038
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.343392848968506,
+      "learning_rate": 4.3648162833729844e-05,
+      "loss": 1.2768,
+      "step": 24039
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.066213607788086,
+      "learning_rate": 4.364789849325932e-05,
+      "loss": 1.4388,
+      "step": 24040
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8079103231430054,
+      "learning_rate": 4.36476341527888e-05,
+      "loss": 1.301,
+      "step": 24041
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.3554675579071045,
+      "learning_rate": 4.3647369812318264e-05,
+      "loss": 1.4903,
+      "step": 24042
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8589836359024048,
+      "learning_rate": 4.3647105471847745e-05,
+      "loss": 1.2721,
+      "step": 24043
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.707263469696045,
+      "learning_rate": 4.364684113137721e-05,
+      "loss": 1.18,
+      "step": 24044
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.6347070932388306,
+      "learning_rate": 4.364657679090669e-05,
+      "loss": 1.5401,
+      "step": 24045
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.981135368347168,
+      "learning_rate": 4.3646312450436165e-05,
+      "loss": 1.2822,
+      "step": 24046
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4153871536254883,
+      "learning_rate": 4.364604810996564e-05,
+      "loss": 1.2833,
+      "step": 24047
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.078179359436035,
+      "learning_rate": 4.364578376949511e-05,
+      "loss": 1.2927,
+      "step": 24048
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1611456871032715,
+      "learning_rate": 4.3645519429024586e-05,
+      "loss": 1.2458,
+      "step": 24049
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9188640117645264,
+      "learning_rate": 4.364525508855406e-05,
+      "loss": 1.3283,
+      "step": 24050
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.44899320602417,
+      "learning_rate": 4.364499074808353e-05,
+      "loss": 1.3549,
+      "step": 24051
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.7797534465789795,
+      "learning_rate": 4.364472640761301e-05,
+      "loss": 1.5225,
+      "step": 24052
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7678070068359375,
+      "learning_rate": 4.364446206714248e-05,
+      "loss": 1.1369,
+      "step": 24053
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.867377758026123,
+      "learning_rate": 4.3644197726671954e-05,
+      "loss": 1.4108,
+      "step": 24054
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5836780071258545,
+      "learning_rate": 4.364393338620143e-05,
+      "loss": 2.0357,
+      "step": 24055
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.784430742263794,
+      "learning_rate": 4.364366904573091e-05,
+      "loss": 1.4184,
+      "step": 24056
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9360934495925903,
+      "learning_rate": 4.3643404705260375e-05,
+      "loss": 1.1245,
+      "step": 24057
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0887155532836914,
+      "learning_rate": 4.3643140364789855e-05,
+      "loss": 1.2748,
+      "step": 24058
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.700907826423645,
+      "learning_rate": 4.364287602431932e-05,
+      "loss": 1.1168,
+      "step": 24059
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0826847553253174,
+      "learning_rate": 4.36426116838488e-05,
+      "loss": 1.3581,
+      "step": 24060
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.219259738922119,
+      "learning_rate": 4.3642347343378276e-05,
+      "loss": 1.2785,
+      "step": 24061
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.489417552947998,
+      "learning_rate": 4.364208300290775e-05,
+      "loss": 1.288,
+      "step": 24062
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9634044170379639,
+      "learning_rate": 4.364181866243722e-05,
+      "loss": 1.3596,
+      "step": 24063
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0786073207855225,
+      "learning_rate": 4.3641554321966696e-05,
+      "loss": 1.4283,
+      "step": 24064
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.2524659633636475,
+      "learning_rate": 4.364128998149617e-05,
+      "loss": 1.6408,
+      "step": 24065
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.257197856903076,
+      "learning_rate": 4.3641025641025643e-05,
+      "loss": 1.2331,
+      "step": 24066
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.07008695602417,
+      "learning_rate": 4.364076130055512e-05,
+      "loss": 1.3626,
+      "step": 24067
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.015451192855835,
+      "learning_rate": 4.364049696008459e-05,
+      "loss": 1.2518,
+      "step": 24068
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.7741987705230713,
+      "learning_rate": 4.3640232619614064e-05,
+      "loss": 1.245,
+      "step": 24069
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.988173246383667,
+      "learning_rate": 4.363996827914354e-05,
+      "loss": 1.2706,
+      "step": 24070
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7607812881469727,
+      "learning_rate": 4.363970393867302e-05,
+      "loss": 1.2791,
+      "step": 24071
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.9578332901000977,
+      "learning_rate": 4.3639439598202485e-05,
+      "loss": 1.5938,
+      "step": 24072
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.638361930847168,
+      "learning_rate": 4.3639175257731965e-05,
+      "loss": 1.2923,
+      "step": 24073
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.00559139251709,
+      "learning_rate": 4.363891091726143e-05,
+      "loss": 1.3304,
+      "step": 24074
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.528350830078125,
+      "learning_rate": 4.363864657679091e-05,
+      "loss": 1.3904,
+      "step": 24075
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.4879857301712036,
+      "learning_rate": 4.3638382236320386e-05,
+      "loss": 1.1319,
+      "step": 24076
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5284959077835083,
+      "learning_rate": 4.363811789584985e-05,
+      "loss": 1.1802,
+      "step": 24077
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.826052188873291,
+      "learning_rate": 4.363785355537933e-05,
+      "loss": 1.2804,
+      "step": 24078
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.752140760421753,
+      "learning_rate": 4.36375892149088e-05,
+      "loss": 1.2654,
+      "step": 24079
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.155555248260498,
+      "learning_rate": 4.363732487443828e-05,
+      "loss": 1.3634,
+      "step": 24080
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.0987229347229004,
+      "learning_rate": 4.3637060533967754e-05,
+      "loss": 1.3869,
+      "step": 24081
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5546965599060059,
+      "learning_rate": 4.363679619349723e-05,
+      "loss": 2.1921,
+      "step": 24082
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1574370861053467,
+      "learning_rate": 4.36365318530267e-05,
+      "loss": 1.3533,
+      "step": 24083
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.990655541419983,
+      "learning_rate": 4.3636267512556175e-05,
+      "loss": 1.2914,
+      "step": 24084
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8969342708587646,
+      "learning_rate": 4.363600317208565e-05,
+      "loss": 1.1869,
+      "step": 24085
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9508394002914429,
+      "learning_rate": 4.363573883161512e-05,
+      "loss": 1.283,
+      "step": 24086
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.2328176498413086,
+      "learning_rate": 4.3635474491144595e-05,
+      "loss": 1.1748,
+      "step": 24087
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.069568157196045,
+      "learning_rate": 4.363521015067407e-05,
+      "loss": 1.3994,
+      "step": 24088
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9863150119781494,
+      "learning_rate": 4.363494581020354e-05,
+      "loss": 1.2406,
+      "step": 24089
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.357063055038452,
+      "learning_rate": 4.3634681469733016e-05,
+      "loss": 1.4265,
+      "step": 24090
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8880773782730103,
+      "learning_rate": 4.3634417129262496e-05,
+      "loss": 1.2709,
+      "step": 24091
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9641509056091309,
+      "learning_rate": 4.363415278879196e-05,
+      "loss": 1.405,
+      "step": 24092
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.913628578186035,
+      "learning_rate": 4.3633888448321443e-05,
+      "loss": 1.3802,
+      "step": 24093
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.348646640777588,
+      "learning_rate": 4.363362410785091e-05,
+      "loss": 1.4543,
+      "step": 24094
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4635865688323975,
+      "learning_rate": 4.363335976738039e-05,
+      "loss": 1.37,
+      "step": 24095
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.60921573638916,
+      "learning_rate": 4.3633095426909864e-05,
+      "loss": 1.4594,
+      "step": 24096
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7950935363769531,
+      "learning_rate": 4.363283108643934e-05,
+      "loss": 1.1498,
+      "step": 24097
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.930198907852173,
+      "learning_rate": 4.363256674596881e-05,
+      "loss": 1.3441,
+      "step": 24098
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.897432565689087,
+      "learning_rate": 4.3632302405498285e-05,
+      "loss": 1.3891,
+      "step": 24099
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4600462913513184,
+      "learning_rate": 4.363203806502776e-05,
+      "loss": 1.4313,
+      "step": 24100
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.5410945415496826,
+      "learning_rate": 4.363177372455723e-05,
+      "loss": 1.6211,
+      "step": 24101
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9527209997177124,
+      "learning_rate": 4.3631509384086706e-05,
+      "loss": 1.2745,
+      "step": 24102
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.924865961074829,
+      "learning_rate": 4.363124504361618e-05,
+      "loss": 1.2551,
+      "step": 24103
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7497022151947021,
+      "learning_rate": 4.363098070314565e-05,
+      "loss": 1.3284,
+      "step": 24104
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.018218994140625,
+      "learning_rate": 4.3630716362675126e-05,
+      "loss": 1.234,
+      "step": 24105
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.1200077533721924,
+      "learning_rate": 4.363045202220461e-05,
+      "loss": 1.2681,
+      "step": 24106
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.890669584274292,
+      "learning_rate": 4.3630187681734073e-05,
+      "loss": 1.2976,
+      "step": 24107
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.91997492313385,
+      "learning_rate": 4.3629923341263554e-05,
+      "loss": 1.2471,
+      "step": 24108
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.3097612857818604,
+      "learning_rate": 4.362965900079302e-05,
+      "loss": 1.3897,
+      "step": 24109
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0682764053344727,
+      "learning_rate": 4.36293946603225e-05,
+      "loss": 1.355,
+      "step": 24110
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.173208475112915,
+      "learning_rate": 4.3629130319851975e-05,
+      "loss": 1.2468,
+      "step": 24111
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.940904378890991,
+      "learning_rate": 4.362886597938145e-05,
+      "loss": 1.3352,
+      "step": 24112
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0183279514312744,
+      "learning_rate": 4.362860163891092e-05,
+      "loss": 1.242,
+      "step": 24113
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.3408544063568115,
+      "learning_rate": 4.362833729844039e-05,
+      "loss": 1.3674,
+      "step": 24114
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.3215131759643555,
+      "learning_rate": 4.362807295796987e-05,
+      "loss": 1.3822,
+      "step": 24115
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0819976329803467,
+      "learning_rate": 4.362780861749934e-05,
+      "loss": 1.1691,
+      "step": 24116
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.0899949073791504,
+      "learning_rate": 4.3627544277028816e-05,
+      "loss": 1.4957,
+      "step": 24117
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9811722040176392,
+      "learning_rate": 4.362727993655829e-05,
+      "loss": 1.3179,
+      "step": 24118
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.036081552505493,
+      "learning_rate": 4.362701559608776e-05,
+      "loss": 1.305,
+      "step": 24119
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.015854597091675,
+      "learning_rate": 4.362675125561724e-05,
+      "loss": 1.2815,
+      "step": 24120
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2407305240631104,
+      "learning_rate": 4.362648691514672e-05,
+      "loss": 1.3744,
+      "step": 24121
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.281130313873291,
+      "learning_rate": 4.3626222574676184e-05,
+      "loss": 1.387,
+      "step": 24122
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.060736656188965,
+      "learning_rate": 4.362595823420566e-05,
+      "loss": 1.2493,
+      "step": 24123
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0577094554901123,
+      "learning_rate": 4.362569389373513e-05,
+      "loss": 1.7191,
+      "step": 24124
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9412643909454346,
+      "learning_rate": 4.3625429553264605e-05,
+      "loss": 1.1667,
+      "step": 24125
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9555549621582031,
+      "learning_rate": 4.3625165212794085e-05,
+      "loss": 1.2491,
+      "step": 24126
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.878251552581787,
+      "learning_rate": 4.362490087232355e-05,
+      "loss": 1.2614,
+      "step": 24127
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.035203218460083,
+      "learning_rate": 4.362463653185303e-05,
+      "loss": 1.3241,
+      "step": 24128
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8476096391677856,
+      "learning_rate": 4.36243721913825e-05,
+      "loss": 1.2834,
+      "step": 24129
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2812907695770264,
+      "learning_rate": 4.362410785091198e-05,
+      "loss": 1.2585,
+      "step": 24130
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6764277219772339,
+      "learning_rate": 4.362384351044145e-05,
+      "loss": 1.3015,
+      "step": 24131
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6941959857940674,
+      "learning_rate": 4.3623579169970926e-05,
+      "loss": 1.1858,
+      "step": 24132
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.056281328201294,
+      "learning_rate": 4.36233148295004e-05,
+      "loss": 1.3491,
+      "step": 24133
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9348559379577637,
+      "learning_rate": 4.3623050489029873e-05,
+      "loss": 1.3365,
+      "step": 24134
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.372354030609131,
+      "learning_rate": 4.362278614855935e-05,
+      "loss": 1.462,
+      "step": 24135
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1263022422790527,
+      "learning_rate": 4.362252180808882e-05,
+      "loss": 1.3062,
+      "step": 24136
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.5402233600616455,
+      "learning_rate": 4.3622257467618294e-05,
+      "loss": 1.112,
+      "step": 24137
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7124879360198975,
+      "learning_rate": 4.362199312714777e-05,
+      "loss": 1.3012,
+      "step": 24138
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.4180452823638916,
+      "learning_rate": 4.362172878667724e-05,
+      "loss": 1.494,
+      "step": 24139
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.300935745239258,
+      "learning_rate": 4.3621464446206715e-05,
+      "loss": 1.3287,
+      "step": 24140
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.4965977668762207,
+      "learning_rate": 4.3621200105736195e-05,
+      "loss": 1.6404,
+      "step": 24141
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.158581256866455,
+      "learning_rate": 4.362093576526566e-05,
+      "loss": 1.3055,
+      "step": 24142
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9917094707489014,
+      "learning_rate": 4.362067142479514e-05,
+      "loss": 1.2554,
+      "step": 24143
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7362682819366455,
+      "learning_rate": 4.362040708432461e-05,
+      "loss": 1.1421,
+      "step": 24144
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9698386192321777,
+      "learning_rate": 4.362014274385409e-05,
+      "loss": 1.3065,
+      "step": 24145
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1217904090881348,
+      "learning_rate": 4.361987840338356e-05,
+      "loss": 1.2342,
+      "step": 24146
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.637583017349243,
+      "learning_rate": 4.361961406291304e-05,
+      "loss": 1.2452,
+      "step": 24147
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.6826670169830322,
+      "learning_rate": 4.361934972244251e-05,
+      "loss": 1.4971,
+      "step": 24148
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.9998884201049805,
+      "learning_rate": 4.361908538197198e-05,
+      "loss": 2.5288,
+      "step": 24149
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.041038990020752,
+      "learning_rate": 4.361882104150146e-05,
+      "loss": 1.3489,
+      "step": 24150
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.6288201808929443,
+      "learning_rate": 4.361855670103093e-05,
+      "loss": 1.2862,
+      "step": 24151
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0967698097229004,
+      "learning_rate": 4.3618292360560404e-05,
+      "loss": 1.2667,
+      "step": 24152
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9219970703125,
+      "learning_rate": 4.361802802008988e-05,
+      "loss": 1.2554,
+      "step": 24153
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9183058738708496,
+      "learning_rate": 4.361776367961935e-05,
+      "loss": 1.9315,
+      "step": 24154
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2436044216156006,
+      "learning_rate": 4.3617499339148825e-05,
+      "loss": 1.3984,
+      "step": 24155
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.807883381843567,
+      "learning_rate": 4.3617234998678306e-05,
+      "loss": 1.2662,
+      "step": 24156
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.087411880493164,
+      "learning_rate": 4.361697065820777e-05,
+      "loss": 1.2212,
+      "step": 24157
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9876210689544678,
+      "learning_rate": 4.3616706317737246e-05,
+      "loss": 1.3983,
+      "step": 24158
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7223237752914429,
+      "learning_rate": 4.361644197726672e-05,
+      "loss": 1.3182,
+      "step": 24159
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9352084398269653,
+      "learning_rate": 4.361617763679619e-05,
+      "loss": 1.3619,
+      "step": 24160
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.886265754699707,
+      "learning_rate": 4.361591329632567e-05,
+      "loss": 1.117,
+      "step": 24161
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.668047547340393,
+      "learning_rate": 4.361564895585514e-05,
+      "loss": 1.102,
+      "step": 24162
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.295334577560425,
+      "learning_rate": 4.361538461538462e-05,
+      "loss": 1.2263,
+      "step": 24163
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2405757904052734,
+      "learning_rate": 4.361512027491409e-05,
+      "loss": 1.2394,
+      "step": 24164
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6741197109222412,
+      "learning_rate": 4.361485593444357e-05,
+      "loss": 1.2864,
+      "step": 24165
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0658223628997803,
+      "learning_rate": 4.361459159397304e-05,
+      "loss": 1.1968,
+      "step": 24166
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2964534759521484,
+      "learning_rate": 4.3614327253502515e-05,
+      "loss": 1.4009,
+      "step": 24167
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2501232624053955,
+      "learning_rate": 4.361406291303199e-05,
+      "loss": 1.276,
+      "step": 24168
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.134052276611328,
+      "learning_rate": 4.361379857256146e-05,
+      "loss": 1.2878,
+      "step": 24169
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8665227890014648,
+      "learning_rate": 4.3613534232090936e-05,
+      "loss": 1.0694,
+      "step": 24170
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9783687591552734,
+      "learning_rate": 4.361326989162041e-05,
+      "loss": 1.292,
+      "step": 24171
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1385161876678467,
+      "learning_rate": 4.361300555114988e-05,
+      "loss": 1.4594,
+      "step": 24172
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0914063453674316,
+      "learning_rate": 4.3612741210679356e-05,
+      "loss": 1.2994,
+      "step": 24173
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.065314769744873,
+      "learning_rate": 4.361247687020883e-05,
+      "loss": 1.396,
+      "step": 24174
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.867591381072998,
+      "learning_rate": 4.36122125297383e-05,
+      "loss": 1.413,
+      "step": 24175
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8212921619415283,
+      "learning_rate": 4.3611948189267784e-05,
+      "loss": 1.317,
+      "step": 24176
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8264126777648926,
+      "learning_rate": 4.361168384879725e-05,
+      "loss": 1.2231,
+      "step": 24177
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0973870754241943,
+      "learning_rate": 4.361141950832673e-05,
+      "loss": 1.3518,
+      "step": 24178
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.085472822189331,
+      "learning_rate": 4.36111551678562e-05,
+      "loss": 1.4138,
+      "step": 24179
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6862428188323975,
+      "learning_rate": 4.361089082738568e-05,
+      "loss": 1.134,
+      "step": 24180
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7682639360427856,
+      "learning_rate": 4.361062648691515e-05,
+      "loss": 1.2324,
+      "step": 24181
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.184804916381836,
+      "learning_rate": 4.3610362146444625e-05,
+      "loss": 1.3347,
+      "step": 24182
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.750144124031067,
+      "learning_rate": 4.36100978059741e-05,
+      "loss": 1.196,
+      "step": 24183
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.018477439880371,
+      "learning_rate": 4.360983346550357e-05,
+      "loss": 1.34,
+      "step": 24184
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9916250705718994,
+      "learning_rate": 4.3609569125033046e-05,
+      "loss": 1.1447,
+      "step": 24185
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.4863903522491455,
+      "learning_rate": 4.360930478456252e-05,
+      "loss": 1.2243,
+      "step": 24186
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.042685031890869,
+      "learning_rate": 4.360904044409199e-05,
+      "loss": 1.2439,
+      "step": 24187
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8535665273666382,
+      "learning_rate": 4.3608776103621467e-05,
+      "loss": 1.2982,
+      "step": 24188
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.6152803897857666,
+      "learning_rate": 4.360851176315094e-05,
+      "loss": 1.3826,
+      "step": 24189
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.7509093284606934,
+      "learning_rate": 4.3608247422680414e-05,
+      "loss": 1.4549,
+      "step": 24190
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8144303560256958,
+      "learning_rate": 4.3607983082209894e-05,
+      "loss": 1.2231,
+      "step": 24191
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.56676983833313,
+      "learning_rate": 4.360771874173936e-05,
+      "loss": 1.4283,
+      "step": 24192
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.036958694458008,
+      "learning_rate": 4.360745440126884e-05,
+      "loss": 1.3107,
+      "step": 24193
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.696800470352173,
+      "learning_rate": 4.360719006079831e-05,
+      "loss": 1.4249,
+      "step": 24194
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6943377256393433,
+      "learning_rate": 4.360692572032778e-05,
+      "loss": 1.2707,
+      "step": 24195
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.231680154800415,
+      "learning_rate": 4.360666137985726e-05,
+      "loss": 1.3936,
+      "step": 24196
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.177186965942383,
+      "learning_rate": 4.360639703938673e-05,
+      "loss": 1.3644,
+      "step": 24197
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8101487159729004,
+      "learning_rate": 4.360613269891621e-05,
+      "loss": 1.2932,
+      "step": 24198
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.235703945159912,
+      "learning_rate": 4.3605868358445676e-05,
+      "loss": 1.3279,
+      "step": 24199
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9657498598098755,
+      "learning_rate": 4.3605604017975156e-05,
+      "loss": 1.3554,
+      "step": 24200
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2888998985290527,
+      "learning_rate": 4.360533967750463e-05,
+      "loss": 1.2851,
+      "step": 24201
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.4245870113372803,
+      "learning_rate": 4.36050753370341e-05,
+      "loss": 1.3898,
+      "step": 24202
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.208731174468994,
+      "learning_rate": 4.360481099656358e-05,
+      "loss": 1.4348,
+      "step": 24203
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.6601450443267822,
+      "learning_rate": 4.360454665609305e-05,
+      "loss": 1.4269,
+      "step": 24204
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8866349458694458,
+      "learning_rate": 4.3604282315622524e-05,
+      "loss": 1.284,
+      "step": 24205
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.645636558532715,
+      "learning_rate": 4.3604017975152e-05,
+      "loss": 1.4305,
+      "step": 24206
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8866620063781738,
+      "learning_rate": 4.360375363468147e-05,
+      "loss": 1.154,
+      "step": 24207
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9238317012786865,
+      "learning_rate": 4.3603489294210945e-05,
+      "loss": 1.3092,
+      "step": 24208
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0815181732177734,
+      "learning_rate": 4.360322495374042e-05,
+      "loss": 1.1292,
+      "step": 24209
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.451448678970337,
+      "learning_rate": 4.360296061326989e-05,
+      "loss": 1.2356,
+      "step": 24210
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.059947967529297,
+      "learning_rate": 4.360269627279937e-05,
+      "loss": 1.3146,
+      "step": 24211
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6290411949157715,
+      "learning_rate": 4.360243193232884e-05,
+      "loss": 1.2155,
+      "step": 24212
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.177109956741333,
+      "learning_rate": 4.360216759185832e-05,
+      "loss": 1.3737,
+      "step": 24213
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.6072897911071777,
+      "learning_rate": 4.3601903251387786e-05,
+      "loss": 1.5079,
+      "step": 24214
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.01424241065979,
+      "learning_rate": 4.3601638910917267e-05,
+      "loss": 1.3436,
+      "step": 24215
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9888343811035156,
+      "learning_rate": 4.360137457044674e-05,
+      "loss": 1.1786,
+      "step": 24216
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.320833206176758,
+      "learning_rate": 4.3601110229976214e-05,
+      "loss": 1.4798,
+      "step": 24217
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.6176586151123047,
+      "learning_rate": 4.360084588950569e-05,
+      "loss": 1.3484,
+      "step": 24218
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.088789463043213,
+      "learning_rate": 4.360058154903516e-05,
+      "loss": 1.2376,
+      "step": 24219
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1855969429016113,
+      "learning_rate": 4.3600317208564634e-05,
+      "loss": 1.4939,
+      "step": 24220
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.126493453979492,
+      "learning_rate": 4.360005286809411e-05,
+      "loss": 1.4194,
+      "step": 24221
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.644493818283081,
+      "learning_rate": 4.359978852762358e-05,
+      "loss": 1.1827,
+      "step": 24222
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.9025216102600098,
+      "learning_rate": 4.3599524187153055e-05,
+      "loss": 1.5495,
+      "step": 24223
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.357558250427246,
+      "learning_rate": 4.359925984668253e-05,
+      "loss": 1.2774,
+      "step": 24224
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.396829605102539,
+      "learning_rate": 4.3598995506212e-05,
+      "loss": 1.3653,
+      "step": 24225
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.588536262512207,
+      "learning_rate": 4.359873116574148e-05,
+      "loss": 1.5304,
+      "step": 24226
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.3425095081329346,
+      "learning_rate": 4.359846682527095e-05,
+      "loss": 1.3984,
+      "step": 24227
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1704578399658203,
+      "learning_rate": 4.359820248480043e-05,
+      "loss": 1.3171,
+      "step": 24228
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.4183294773101807,
+      "learning_rate": 4.3597938144329897e-05,
+      "loss": 1.234,
+      "step": 24229
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.110928773880005,
+      "learning_rate": 4.359767380385938e-05,
+      "loss": 1.3953,
+      "step": 24230
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0841665267944336,
+      "learning_rate": 4.359740946338885e-05,
+      "loss": 1.2193,
+      "step": 24231
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.2937896251678467,
+      "learning_rate": 4.359714512291832e-05,
+      "loss": 1.5841,
+      "step": 24232
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8356900215148926,
+      "learning_rate": 4.35968807824478e-05,
+      "loss": 1.3512,
+      "step": 24233
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.569369316101074,
+      "learning_rate": 4.3596616441977264e-05,
+      "loss": 1.4237,
+      "step": 24234
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6677744388580322,
+      "learning_rate": 4.3596352101506745e-05,
+      "loss": 1.1246,
+      "step": 24235
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.3291678428649902,
+      "learning_rate": 4.359608776103622e-05,
+      "loss": 1.439,
+      "step": 24236
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0179800987243652,
+      "learning_rate": 4.359582342056569e-05,
+      "loss": 1.2213,
+      "step": 24237
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.4659783840179443,
+      "learning_rate": 4.3595559080095165e-05,
+      "loss": 1.8594,
+      "step": 24238
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.89482843875885,
+      "learning_rate": 4.359529473962464e-05,
+      "loss": 1.2994,
+      "step": 24239
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.1373133659362793,
+      "learning_rate": 4.359503039915411e-05,
+      "loss": 1.2591,
+      "step": 24240
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1588540077209473,
+      "learning_rate": 4.3594766058683586e-05,
+      "loss": 1.4105,
+      "step": 24241
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.5071911811828613,
+      "learning_rate": 4.359450171821306e-05,
+      "loss": 1.9402,
+      "step": 24242
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.770173192024231,
+      "learning_rate": 4.359423737774253e-05,
+      "loss": 1.5304,
+      "step": 24243
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.080925941467285,
+      "learning_rate": 4.359397303727201e-05,
+      "loss": 1.3138,
+      "step": 24244
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9439624547958374,
+      "learning_rate": 4.359370869680148e-05,
+      "loss": 1.291,
+      "step": 24245
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.5692001581192017,
+      "learning_rate": 4.3593444356330954e-05,
+      "loss": 1.1204,
+      "step": 24246
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.420154571533203,
+      "learning_rate": 4.359318001586043e-05,
+      "loss": 1.3065,
+      "step": 24247
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1206588745117188,
+      "learning_rate": 4.359291567538991e-05,
+      "loss": 1.371,
+      "step": 24248
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9021573066711426,
+      "learning_rate": 4.3592651334919375e-05,
+      "loss": 1.3456,
+      "step": 24249
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8436174392700195,
+      "learning_rate": 4.3592386994448855e-05,
+      "loss": 1.3673,
+      "step": 24250
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8333789110183716,
+      "learning_rate": 4.359212265397832e-05,
+      "loss": 1.6677,
+      "step": 24251
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.072348117828369,
+      "learning_rate": 4.35918583135078e-05,
+      "loss": 1.2689,
+      "step": 24252
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8631458282470703,
+      "learning_rate": 4.3591593973037276e-05,
+      "loss": 1.3899,
+      "step": 24253
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1402580738067627,
+      "learning_rate": 4.359132963256675e-05,
+      "loss": 1.3298,
+      "step": 24254
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.168053150177002,
+      "learning_rate": 4.359106529209622e-05,
+      "loss": 1.2814,
+      "step": 24255
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.155560255050659,
+      "learning_rate": 4.3590800951625697e-05,
+      "loss": 1.3426,
+      "step": 24256
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9213035106658936,
+      "learning_rate": 4.359053661115517e-05,
+      "loss": 1.312,
+      "step": 24257
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.3188769817352295,
+      "learning_rate": 4.3590272270684644e-05,
+      "loss": 1.2426,
+      "step": 24258
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6500040292739868,
+      "learning_rate": 4.359000793021412e-05,
+      "loss": 1.2995,
+      "step": 24259
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.744823694229126,
+      "learning_rate": 4.358974358974359e-05,
+      "loss": 1.486,
+      "step": 24260
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.290142774581909,
+      "learning_rate": 4.3589479249273064e-05,
+      "loss": 1.3097,
+      "step": 24261
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.418731212615967,
+      "learning_rate": 4.358921490880254e-05,
+      "loss": 1.3358,
+      "step": 24262
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.101454973220825,
+      "learning_rate": 4.358895056833202e-05,
+      "loss": 1.4207,
+      "step": 24263
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2246592044830322,
+      "learning_rate": 4.3588686227861485e-05,
+      "loss": 1.3431,
+      "step": 24264
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.5789480209350586,
+      "learning_rate": 4.3588421887390965e-05,
+      "loss": 1.184,
+      "step": 24265
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1233763694763184,
+      "learning_rate": 4.358815754692043e-05,
+      "loss": 1.3434,
+      "step": 24266
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.6999459266662598,
+      "learning_rate": 4.3587893206449906e-05,
+      "loss": 2.1401,
+      "step": 24267
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.392193555831909,
+      "learning_rate": 4.3587628865979386e-05,
+      "loss": 1.2208,
+      "step": 24268
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1798899173736572,
+      "learning_rate": 4.358736452550885e-05,
+      "loss": 1.4848,
+      "step": 24269
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.4500808715820312,
+      "learning_rate": 4.358710018503833e-05,
+      "loss": 1.2298,
+      "step": 24270
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7040575742721558,
+      "learning_rate": 4.35868358445678e-05,
+      "loss": 1.257,
+      "step": 24271
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.116536855697632,
+      "learning_rate": 4.358657150409728e-05,
+      "loss": 1.347,
+      "step": 24272
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.4661099910736084,
+      "learning_rate": 4.3586307163626754e-05,
+      "loss": 1.2794,
+      "step": 24273
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.199390172958374,
+      "learning_rate": 4.358604282315623e-05,
+      "loss": 1.4033,
+      "step": 24274
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.132808208465576,
+      "learning_rate": 4.35857784826857e-05,
+      "loss": 1.403,
+      "step": 24275
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.5291165113449097,
+      "learning_rate": 4.3585514142215175e-05,
+      "loss": 1.2158,
+      "step": 24276
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8600118160247803,
+      "learning_rate": 4.358524980174465e-05,
+      "loss": 1.2675,
+      "step": 24277
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.5940423011779785,
+      "learning_rate": 4.358498546127412e-05,
+      "loss": 1.3317,
+      "step": 24278
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6847118139266968,
+      "learning_rate": 4.3584721120803595e-05,
+      "loss": 1.287,
+      "step": 24279
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.068387985229492,
+      "learning_rate": 4.358445678033307e-05,
+      "loss": 1.2364,
+      "step": 24280
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.267674684524536,
+      "learning_rate": 4.358419243986254e-05,
+      "loss": 1.2655,
+      "step": 24281
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.705280303955078,
+      "learning_rate": 4.3583928099392016e-05,
+      "loss": 1.4424,
+      "step": 24282
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.568426489830017,
+      "learning_rate": 4.3583663758921496e-05,
+      "loss": 1.5609,
+      "step": 24283
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0219051837921143,
+      "learning_rate": 4.358339941845096e-05,
+      "loss": 1.2012,
+      "step": 24284
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.999040961265564,
+      "learning_rate": 4.3583135077980444e-05,
+      "loss": 1.3229,
+      "step": 24285
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.732266902923584,
+      "learning_rate": 4.358287073750991e-05,
+      "loss": 1.2759,
+      "step": 24286
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8804223537445068,
+      "learning_rate": 4.358260639703939e-05,
+      "loss": 1.412,
+      "step": 24287
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6615655422210693,
+      "learning_rate": 4.3582342056568864e-05,
+      "loss": 1.2261,
+      "step": 24288
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1356446743011475,
+      "learning_rate": 4.358207771609834e-05,
+      "loss": 1.2894,
+      "step": 24289
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.710292100906372,
+      "learning_rate": 4.358181337562781e-05,
+      "loss": 2.6062,
+      "step": 24290
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.062131881713867,
+      "learning_rate": 4.3581549035157285e-05,
+      "loss": 1.1525,
+      "step": 24291
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.840783953666687,
+      "learning_rate": 4.358128469468676e-05,
+      "loss": 1.4639,
+      "step": 24292
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.733158826828003,
+      "learning_rate": 4.358102035421623e-05,
+      "loss": 1.1714,
+      "step": 24293
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.2296805381774902,
+      "learning_rate": 4.3580756013745706e-05,
+      "loss": 1.3727,
+      "step": 24294
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7294058799743652,
+      "learning_rate": 4.358049167327518e-05,
+      "loss": 1.2601,
+      "step": 24295
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0049586296081543,
+      "learning_rate": 4.358022733280465e-05,
+      "loss": 1.4186,
+      "step": 24296
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2818987369537354,
+      "learning_rate": 4.3579962992334126e-05,
+      "loss": 1.4931,
+      "step": 24297
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.052009344100952,
+      "learning_rate": 4.357969865186361e-05,
+      "loss": 1.2218,
+      "step": 24298
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9351791143417358,
+      "learning_rate": 4.3579434311393074e-05,
+      "loss": 1.2009,
+      "step": 24299
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.133019208908081,
+      "learning_rate": 4.3579169970922554e-05,
+      "loss": 1.3137,
+      "step": 24300
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.729780673980713,
+      "learning_rate": 4.357890563045202e-05,
+      "loss": 1.1825,
+      "step": 24301
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.5790899991989136,
+      "learning_rate": 4.35786412899815e-05,
+      "loss": 1.2459,
+      "step": 24302
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9765719175338745,
+      "learning_rate": 4.3578376949510975e-05,
+      "loss": 1.3061,
+      "step": 24303
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7187646627426147,
+      "learning_rate": 4.357811260904044e-05,
+      "loss": 1.2608,
+      "step": 24304
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.445005416870117,
+      "learning_rate": 4.357784826856992e-05,
+      "loss": 1.5881,
+      "step": 24305
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2998528480529785,
+      "learning_rate": 4.357758392809939e-05,
+      "loss": 1.3027,
+      "step": 24306
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.572322368621826,
+      "learning_rate": 4.357731958762887e-05,
+      "loss": 1.3547,
+      "step": 24307
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6694995164871216,
+      "learning_rate": 4.357705524715834e-05,
+      "loss": 1.0493,
+      "step": 24308
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0575501918792725,
+      "learning_rate": 4.3576790906687816e-05,
+      "loss": 1.2507,
+      "step": 24309
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8637210130691528,
+      "learning_rate": 4.357652656621729e-05,
+      "loss": 1.2153,
+      "step": 24310
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0659549236297607,
+      "learning_rate": 4.357626222574676e-05,
+      "loss": 1.2933,
+      "step": 24311
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8300679922103882,
+      "learning_rate": 4.357599788527624e-05,
+      "loss": 1.3676,
+      "step": 24312
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1155412197113037,
+      "learning_rate": 4.357573354480571e-05,
+      "loss": 1.425,
+      "step": 24313
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2887191772460938,
+      "learning_rate": 4.3575469204335184e-05,
+      "loss": 1.5092,
+      "step": 24314
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.266528844833374,
+      "learning_rate": 4.357520486386466e-05,
+      "loss": 1.3323,
+      "step": 24315
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7064756155014038,
+      "learning_rate": 4.357494052339413e-05,
+      "loss": 1.1866,
+      "step": 24316
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9345433712005615,
+      "learning_rate": 4.3574676182923605e-05,
+      "loss": 1.2864,
+      "step": 24317
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.4962875843048096,
+      "learning_rate": 4.3574411842453085e-05,
+      "loss": 1.3517,
+      "step": 24318
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.074512481689453,
+      "learning_rate": 4.357414750198255e-05,
+      "loss": 1.0194,
+      "step": 24319
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.111682891845703,
+      "learning_rate": 4.357388316151203e-05,
+      "loss": 1.2952,
+      "step": 24320
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3059675693511963,
+      "learning_rate": 4.35736188210415e-05,
+      "loss": 1.3598,
+      "step": 24321
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2072064876556396,
+      "learning_rate": 4.357335448057098e-05,
+      "loss": 1.2635,
+      "step": 24322
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9372773170471191,
+      "learning_rate": 4.357309014010045e-05,
+      "loss": 1.2628,
+      "step": 24323
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2265119552612305,
+      "learning_rate": 4.3572825799629926e-05,
+      "loss": 1.3381,
+      "step": 24324
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.5080044269561768,
+      "learning_rate": 4.35725614591594e-05,
+      "loss": 1.3448,
+      "step": 24325
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6341608762741089,
+      "learning_rate": 4.3572297118688874e-05,
+      "loss": 1.082,
+      "step": 24326
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2428319454193115,
+      "learning_rate": 4.357203277821835e-05,
+      "loss": 1.276,
+      "step": 24327
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.791451096534729,
+      "learning_rate": 4.357176843774782e-05,
+      "loss": 1.3218,
+      "step": 24328
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7520592212677002,
+      "learning_rate": 4.3571504097277294e-05,
+      "loss": 1.124,
+      "step": 24329
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9681893587112427,
+      "learning_rate": 4.357123975680677e-05,
+      "loss": 1.2627,
+      "step": 24330
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.7838692665100098,
+      "learning_rate": 4.357097541633624e-05,
+      "loss": 1.4953,
+      "step": 24331
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9627349376678467,
+      "learning_rate": 4.3570711075865715e-05,
+      "loss": 1.3115,
+      "step": 24332
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.044466018676758,
+      "learning_rate": 4.3570446735395195e-05,
+      "loss": 1.1779,
+      "step": 24333
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.4649441242218018,
+      "learning_rate": 4.357018239492466e-05,
+      "loss": 1.3684,
+      "step": 24334
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6713322401046753,
+      "learning_rate": 4.356991805445414e-05,
+      "loss": 1.258,
+      "step": 24335
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1395230293273926,
+      "learning_rate": 4.356965371398361e-05,
+      "loss": 1.3091,
+      "step": 24336
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.097344398498535,
+      "learning_rate": 4.356938937351309e-05,
+      "loss": 1.3406,
+      "step": 24337
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6388691663742065,
+      "learning_rate": 4.356912503304256e-05,
+      "loss": 1.2175,
+      "step": 24338
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7506074905395508,
+      "learning_rate": 4.356886069257203e-05,
+      "loss": 1.0888,
+      "step": 24339
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3924121856689453,
+      "learning_rate": 4.356859635210151e-05,
+      "loss": 1.3668,
+      "step": 24340
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2524232864379883,
+      "learning_rate": 4.356833201163098e-05,
+      "loss": 1.267,
+      "step": 24341
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.953105092048645,
+      "learning_rate": 4.356806767116046e-05,
+      "loss": 1.2496,
+      "step": 24342
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.01460337638855,
+      "learning_rate": 4.356780333068993e-05,
+      "loss": 1.3125,
+      "step": 24343
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2169342041015625,
+      "learning_rate": 4.3567538990219405e-05,
+      "loss": 2.9476,
+      "step": 24344
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0081980228424072,
+      "learning_rate": 4.356727464974888e-05,
+      "loss": 1.9573,
+      "step": 24345
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0443520545959473,
+      "learning_rate": 4.356701030927835e-05,
+      "loss": 1.1982,
+      "step": 24346
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.6764211654663086,
+      "learning_rate": 4.3566745968807825e-05,
+      "loss": 1.2261,
+      "step": 24347
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8939625024795532,
+      "learning_rate": 4.35664816283373e-05,
+      "loss": 1.225,
+      "step": 24348
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1336300373077393,
+      "learning_rate": 4.356621728786677e-05,
+      "loss": 1.3883,
+      "step": 24349
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8706549406051636,
+      "learning_rate": 4.3565952947396246e-05,
+      "loss": 1.2759,
+      "step": 24350
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7549265623092651,
+      "learning_rate": 4.356568860692572e-05,
+      "loss": 1.3704,
+      "step": 24351
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.9911749362945557,
+      "learning_rate": 4.356542426645519e-05,
+      "loss": 1.2118,
+      "step": 24352
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.4192190170288086,
+      "learning_rate": 4.3565159925984674e-05,
+      "loss": 1.2951,
+      "step": 24353
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7060463428497314,
+      "learning_rate": 4.356489558551414e-05,
+      "loss": 1.0936,
+      "step": 24354
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.022674083709717,
+      "learning_rate": 4.356463124504362e-05,
+      "loss": 1.2349,
+      "step": 24355
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9842264652252197,
+      "learning_rate": 4.356436690457309e-05,
+      "loss": 1.3939,
+      "step": 24356
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8720598220825195,
+      "learning_rate": 4.356410256410257e-05,
+      "loss": 1.2743,
+      "step": 24357
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9628292322158813,
+      "learning_rate": 4.356383822363204e-05,
+      "loss": 1.2857,
+      "step": 24358
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1778125762939453,
+      "learning_rate": 4.3563573883161515e-05,
+      "loss": 2.0831,
+      "step": 24359
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3716208934783936,
+      "learning_rate": 4.356330954269099e-05,
+      "loss": 1.2063,
+      "step": 24360
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3144960403442383,
+      "learning_rate": 4.356304520222046e-05,
+      "loss": 1.3601,
+      "step": 24361
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.8665051460266113,
+      "learning_rate": 4.3562780861749936e-05,
+      "loss": 1.5046,
+      "step": 24362
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9274964332580566,
+      "learning_rate": 4.356251652127941e-05,
+      "loss": 1.3317,
+      "step": 24363
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0000429153442383,
+      "learning_rate": 4.356225218080888e-05,
+      "loss": 1.2873,
+      "step": 24364
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8675135374069214,
+      "learning_rate": 4.3561987840338356e-05,
+      "loss": 1.3184,
+      "step": 24365
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7690162658691406,
+      "learning_rate": 4.356172349986783e-05,
+      "loss": 1.3457,
+      "step": 24366
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.847314834594727,
+      "learning_rate": 4.3561459159397304e-05,
+      "loss": 1.6667,
+      "step": 24367
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.4682412147521973,
+      "learning_rate": 4.3561194818926784e-05,
+      "loss": 1.4113,
+      "step": 24368
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9405261278152466,
+      "learning_rate": 4.356093047845625e-05,
+      "loss": 1.3207,
+      "step": 24369
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1988685131073,
+      "learning_rate": 4.356066613798573e-05,
+      "loss": 1.3578,
+      "step": 24370
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.078070163726807,
+      "learning_rate": 4.35604017975152e-05,
+      "loss": 1.4471,
+      "step": 24371
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.217691421508789,
+      "learning_rate": 4.356013745704468e-05,
+      "loss": 1.3138,
+      "step": 24372
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8531475067138672,
+      "learning_rate": 4.355987311657415e-05,
+      "loss": 1.4056,
+      "step": 24373
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.8760807514190674,
+      "learning_rate": 4.3559608776103625e-05,
+      "loss": 1.3516,
+      "step": 24374
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.138327121734619,
+      "learning_rate": 4.35593444356331e-05,
+      "loss": 1.2427,
+      "step": 24375
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.5749964714050293,
+      "learning_rate": 4.3559080095162566e-05,
+      "loss": 1.763,
+      "step": 24376
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.5319015979766846,
+      "learning_rate": 4.3558815754692046e-05,
+      "loss": 1.5468,
+      "step": 24377
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8703703880310059,
+      "learning_rate": 4.355855141422152e-05,
+      "loss": 1.2341,
+      "step": 24378
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7693049907684326,
+      "learning_rate": 4.355828707375099e-05,
+      "loss": 1.2771,
+      "step": 24379
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8820545673370361,
+      "learning_rate": 4.355802273328047e-05,
+      "loss": 1.3721,
+      "step": 24380
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2887890338897705,
+      "learning_rate": 4.355775839280994e-05,
+      "loss": 1.5783,
+      "step": 24381
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9563161134719849,
+      "learning_rate": 4.3557494052339414e-05,
+      "loss": 2.1055,
+      "step": 24382
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.5730140209198,
+      "learning_rate": 4.3557229711868894e-05,
+      "loss": 1.477,
+      "step": 24383
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7375773191452026,
+      "learning_rate": 4.355696537139836e-05,
+      "loss": 1.695,
+      "step": 24384
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.281543254852295,
+      "learning_rate": 4.3556701030927835e-05,
+      "loss": 1.2356,
+      "step": 24385
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6044026613235474,
+      "learning_rate": 4.355643669045731e-05,
+      "loss": 1.2156,
+      "step": 24386
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2949509620666504,
+      "learning_rate": 4.355617234998678e-05,
+      "loss": 1.3465,
+      "step": 24387
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.617161989212036,
+      "learning_rate": 4.355590800951626e-05,
+      "loss": 1.363,
+      "step": 24388
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9069163799285889,
+      "learning_rate": 4.355564366904573e-05,
+      "loss": 1.309,
+      "step": 24389
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.7350051403045654,
+      "learning_rate": 4.355537932857521e-05,
+      "loss": 1.1957,
+      "step": 24390
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3622357845306396,
+      "learning_rate": 4.3555114988104676e-05,
+      "loss": 1.438,
+      "step": 24391
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.426713228225708,
+      "learning_rate": 4.3554850647634156e-05,
+      "loss": 1.4277,
+      "step": 24392
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7734346389770508,
+      "learning_rate": 4.355458630716363e-05,
+      "loss": 1.2417,
+      "step": 24393
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2796096801757812,
+      "learning_rate": 4.3554321966693104e-05,
+      "loss": 1.3317,
+      "step": 24394
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6999479532241821,
+      "learning_rate": 4.355405762622258e-05,
+      "loss": 1.2088,
+      "step": 24395
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.256077289581299,
+      "learning_rate": 4.355379328575205e-05,
+      "loss": 1.6099,
+      "step": 24396
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0336766242980957,
+      "learning_rate": 4.3553528945281524e-05,
+      "loss": 1.2603,
+      "step": 24397
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2151105403900146,
+      "learning_rate": 4.3553264604811e-05,
+      "loss": 1.2993,
+      "step": 24398
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7047598361968994,
+      "learning_rate": 4.355300026434047e-05,
+      "loss": 1.1322,
+      "step": 24399
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.9299042224884033,
+      "learning_rate": 4.3552735923869945e-05,
+      "loss": 1.3598,
+      "step": 24400
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3316032886505127,
+      "learning_rate": 4.355247158339942e-05,
+      "loss": 1.2978,
+      "step": 24401
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8829809427261353,
+      "learning_rate": 4.355220724292889e-05,
+      "loss": 1.2383,
+      "step": 24402
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.792910099029541,
+      "learning_rate": 4.355194290245837e-05,
+      "loss": 1.3993,
+      "step": 24403
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.907945990562439,
+      "learning_rate": 4.355167856198784e-05,
+      "loss": 1.3303,
+      "step": 24404
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2285211086273193,
+      "learning_rate": 4.355141422151732e-05,
+      "loss": 1.4462,
+      "step": 24405
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.274868965148926,
+      "learning_rate": 4.3551149881046786e-05,
+      "loss": 1.3089,
+      "step": 24406
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.4666287899017334,
+      "learning_rate": 4.355088554057627e-05,
+      "loss": 1.3903,
+      "step": 24407
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0438666343688965,
+      "learning_rate": 4.355062120010574e-05,
+      "loss": 1.3467,
+      "step": 24408
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.964605689048767,
+      "learning_rate": 4.3550356859635214e-05,
+      "loss": 1.3186,
+      "step": 24409
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8748905658721924,
+      "learning_rate": 4.355009251916469e-05,
+      "loss": 1.1253,
+      "step": 24410
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3500921726226807,
+      "learning_rate": 4.3549828178694154e-05,
+      "loss": 1.356,
+      "step": 24411
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.2393112182617188,
+      "learning_rate": 4.3549563838223635e-05,
+      "loss": 1.5837,
+      "step": 24412
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.096182346343994,
+      "learning_rate": 4.354929949775311e-05,
+      "loss": 1.2839,
+      "step": 24413
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8718193769454956,
+      "learning_rate": 4.354903515728258e-05,
+      "loss": 1.2382,
+      "step": 24414
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8822952508926392,
+      "learning_rate": 4.3548770816812055e-05,
+      "loss": 1.3475,
+      "step": 24415
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.088365077972412,
+      "learning_rate": 4.354850647634153e-05,
+      "loss": 1.3089,
+      "step": 24416
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7700217962265015,
+      "learning_rate": 4.3548242135871e-05,
+      "loss": 1.3595,
+      "step": 24417
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6311358213424683,
+      "learning_rate": 4.354797779540048e-05,
+      "loss": 1.2554,
+      "step": 24418
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9228620529174805,
+      "learning_rate": 4.354771345492995e-05,
+      "loss": 1.2319,
+      "step": 24419
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.163583278656006,
+      "learning_rate": 4.354744911445943e-05,
+      "loss": 1.1916,
+      "step": 24420
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0622212886810303,
+      "learning_rate": 4.35471847739889e-05,
+      "loss": 1.2563,
+      "step": 24421
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.195441961288452,
+      "learning_rate": 4.354692043351837e-05,
+      "loss": 1.3304,
+      "step": 24422
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3724918365478516,
+      "learning_rate": 4.354665609304785e-05,
+      "loss": 1.3002,
+      "step": 24423
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7793948650360107,
+      "learning_rate": 4.354639175257732e-05,
+      "loss": 1.2053,
+      "step": 24424
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2515578269958496,
+      "learning_rate": 4.35461274121068e-05,
+      "loss": 1.3071,
+      "step": 24425
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8897196054458618,
+      "learning_rate": 4.3545863071636265e-05,
+      "loss": 1.2951,
+      "step": 24426
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7361886501312256,
+      "learning_rate": 4.3545598731165745e-05,
+      "loss": 1.3536,
+      "step": 24427
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.052069664001465,
+      "learning_rate": 4.354533439069522e-05,
+      "loss": 1.3219,
+      "step": 24428
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.459883213043213,
+      "learning_rate": 4.354507005022469e-05,
+      "loss": 1.259,
+      "step": 24429
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.891120433807373,
+      "learning_rate": 4.3544805709754166e-05,
+      "loss": 1.2975,
+      "step": 24430
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.463350296020508,
+      "learning_rate": 4.354454136928364e-05,
+      "loss": 1.46,
+      "step": 24431
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0835001468658447,
+      "learning_rate": 4.354427702881311e-05,
+      "loss": 1.2921,
+      "step": 24432
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8164750337600708,
+      "learning_rate": 4.3544012688342586e-05,
+      "loss": 1.1805,
+      "step": 24433
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7936909198760986,
+      "learning_rate": 4.354374834787206e-05,
+      "loss": 1.2281,
+      "step": 24434
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.533874034881592,
+      "learning_rate": 4.3543484007401533e-05,
+      "loss": 1.534,
+      "step": 24435
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.024766683578491,
+      "learning_rate": 4.354321966693101e-05,
+      "loss": 1.331,
+      "step": 24436
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9484726190567017,
+      "learning_rate": 4.354295532646048e-05,
+      "loss": 1.2568,
+      "step": 24437
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.901157021522522,
+      "learning_rate": 4.354269098598996e-05,
+      "loss": 1.1885,
+      "step": 24438
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.323768377304077,
+      "learning_rate": 4.354242664551943e-05,
+      "loss": 1.4933,
+      "step": 24439
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9739925861358643,
+      "learning_rate": 4.354216230504891e-05,
+      "loss": 1.2877,
+      "step": 24440
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.519502639770508,
+      "learning_rate": 4.3541897964578375e-05,
+      "loss": 1.5005,
+      "step": 24441
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.004598379135132,
+      "learning_rate": 4.3541633624107855e-05,
+      "loss": 1.2685,
+      "step": 24442
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2085459232330322,
+      "learning_rate": 4.354136928363733e-05,
+      "loss": 1.2881,
+      "step": 24443
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1289756298065186,
+      "learning_rate": 4.35411049431668e-05,
+      "loss": 1.2439,
+      "step": 24444
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0780203342437744,
+      "learning_rate": 4.3540840602696276e-05,
+      "loss": 1.3357,
+      "step": 24445
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.341514825820923,
+      "learning_rate": 4.354057626222575e-05,
+      "loss": 1.3266,
+      "step": 24446
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.4728236198425293,
+      "learning_rate": 4.354031192175522e-05,
+      "loss": 1.5456,
+      "step": 24447
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0290684700012207,
+      "learning_rate": 4.35400475812847e-05,
+      "loss": 1.3469,
+      "step": 24448
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9265998601913452,
+      "learning_rate": 4.353978324081417e-05,
+      "loss": 1.2428,
+      "step": 24449
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9145904779434204,
+      "learning_rate": 4.3539518900343644e-05,
+      "loss": 1.216,
+      "step": 24450
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9761369228363037,
+      "learning_rate": 4.353925455987312e-05,
+      "loss": 1.2772,
+      "step": 24451
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.6716034412384033,
+      "learning_rate": 4.353899021940259e-05,
+      "loss": 1.4299,
+      "step": 24452
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2782726287841797,
+      "learning_rate": 4.353872587893207e-05,
+      "loss": 1.064,
+      "step": 24453
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3142035007476807,
+      "learning_rate": 4.353846153846154e-05,
+      "loss": 1.2886,
+      "step": 24454
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9945626258850098,
+      "learning_rate": 4.353819719799102e-05,
+      "loss": 1.3627,
+      "step": 24455
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.896633267402649,
+      "learning_rate": 4.3537932857520485e-05,
+      "loss": 1.3086,
+      "step": 24456
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0388073921203613,
+      "learning_rate": 4.353766851704996e-05,
+      "loss": 1.1458,
+      "step": 24457
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.305931568145752,
+      "learning_rate": 4.353740417657944e-05,
+      "loss": 1.4138,
+      "step": 24458
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.4980061054229736,
+      "learning_rate": 4.3537139836108906e-05,
+      "loss": 1.4433,
+      "step": 24459
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2406089305877686,
+      "learning_rate": 4.3536875495638386e-05,
+      "loss": 1.3104,
+      "step": 24460
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1768505573272705,
+      "learning_rate": 4.353661115516785e-05,
+      "loss": 1.4532,
+      "step": 24461
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.537684917449951,
+      "learning_rate": 4.3536346814697333e-05,
+      "loss": 1.3987,
+      "step": 24462
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9626128673553467,
+      "learning_rate": 4.353608247422681e-05,
+      "loss": 1.3108,
+      "step": 24463
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.4602198600769043,
+      "learning_rate": 4.353581813375628e-05,
+      "loss": 1.3721,
+      "step": 24464
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.4650514125823975,
+      "learning_rate": 4.3535553793285754e-05,
+      "loss": 1.3563,
+      "step": 24465
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8648443222045898,
+      "learning_rate": 4.353528945281523e-05,
+      "loss": 1.3385,
+      "step": 24466
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.598931908607483,
+      "learning_rate": 4.35350251123447e-05,
+      "loss": 1.3118,
+      "step": 24467
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9499421119689941,
+      "learning_rate": 4.3534760771874175e-05,
+      "loss": 1.3445,
+      "step": 24468
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.981252908706665,
+      "learning_rate": 4.353449643140365e-05,
+      "loss": 1.3482,
+      "step": 24469
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.662650465965271,
+      "learning_rate": 4.353423209093312e-05,
+      "loss": 1.0755,
+      "step": 24470
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.747107744216919,
+      "learning_rate": 4.3533967750462596e-05,
+      "loss": 1.1722,
+      "step": 24471
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.727711796760559,
+      "learning_rate": 4.353370340999207e-05,
+      "loss": 1.2287,
+      "step": 24472
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.621610403060913,
+      "learning_rate": 4.353343906952155e-05,
+      "loss": 1.3313,
+      "step": 24473
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8894240856170654,
+      "learning_rate": 4.3533174729051016e-05,
+      "loss": 1.2701,
+      "step": 24474
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3161513805389404,
+      "learning_rate": 4.35329103885805e-05,
+      "loss": 1.3774,
+      "step": 24475
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0661234855651855,
+      "learning_rate": 4.3532646048109963e-05,
+      "loss": 1.3654,
+      "step": 24476
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1401755809783936,
+      "learning_rate": 4.3532381707639444e-05,
+      "loss": 1.2541,
+      "step": 24477
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2237842082977295,
+      "learning_rate": 4.353211736716892e-05,
+      "loss": 1.4213,
+      "step": 24478
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0343830585479736,
+      "learning_rate": 4.353185302669839e-05,
+      "loss": 1.2532,
+      "step": 24479
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7099909782409668,
+      "learning_rate": 4.3531588686227864e-05,
+      "loss": 1.2709,
+      "step": 24480
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9308267831802368,
+      "learning_rate": 4.353132434575734e-05,
+      "loss": 1.2277,
+      "step": 24481
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9448020458221436,
+      "learning_rate": 4.353106000528681e-05,
+      "loss": 1.2682,
+      "step": 24482
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1128246784210205,
+      "learning_rate": 4.3530795664816285e-05,
+      "loss": 1.2734,
+      "step": 24483
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3795619010925293,
+      "learning_rate": 4.353053132434576e-05,
+      "loss": 1.457,
+      "step": 24484
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.423696756362915,
+      "learning_rate": 4.353026698387523e-05,
+      "loss": 2.4805,
+      "step": 24485
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7093322277069092,
+      "learning_rate": 4.3530002643404706e-05,
+      "loss": 1.1843,
+      "step": 24486
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.104041337966919,
+      "learning_rate": 4.352973830293418e-05,
+      "loss": 1.3262,
+      "step": 24487
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.119529962539673,
+      "learning_rate": 4.352947396246366e-05,
+      "loss": 1.2749,
+      "step": 24488
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0146915912628174,
+      "learning_rate": 4.352920962199313e-05,
+      "loss": 1.3053,
+      "step": 24489
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.268190860748291,
+      "learning_rate": 4.352894528152261e-05,
+      "loss": 1.3252,
+      "step": 24490
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.4914413690567017,
+      "learning_rate": 4.3528680941052074e-05,
+      "loss": 1.198,
+      "step": 24491
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.047724962234497,
+      "learning_rate": 4.3528416600581554e-05,
+      "loss": 1.4942,
+      "step": 24492
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.770600438117981,
+      "learning_rate": 4.352815226011103e-05,
+      "loss": 1.4325,
+      "step": 24493
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.40777587890625,
+      "learning_rate": 4.3527887919640494e-05,
+      "loss": 1.3598,
+      "step": 24494
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3791005611419678,
+      "learning_rate": 4.3527623579169975e-05,
+      "loss": 1.3824,
+      "step": 24495
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2959885597229004,
+      "learning_rate": 4.352735923869944e-05,
+      "loss": 1.365,
+      "step": 24496
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1846868991851807,
+      "learning_rate": 4.352709489822892e-05,
+      "loss": 1.3274,
+      "step": 24497
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.6100289821624756,
+      "learning_rate": 4.3526830557758396e-05,
+      "loss": 1.3999,
+      "step": 24498
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.116600513458252,
+      "learning_rate": 4.352656621728787e-05,
+      "loss": 1.204,
+      "step": 24499
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.410717010498047,
+      "learning_rate": 4.352630187681734e-05,
+      "loss": 1.2864,
+      "step": 24500
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.038362503051758,
+      "learning_rate": 4.3526037536346816e-05,
+      "loss": 1.1231,
+      "step": 24501
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.760613203048706,
+      "learning_rate": 4.352577319587629e-05,
+      "loss": 1.2234,
+      "step": 24502
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.7778830528259277,
+      "learning_rate": 4.3525508855405763e-05,
+      "loss": 1.4806,
+      "step": 24503
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.055173635482788,
+      "learning_rate": 4.352524451493524e-05,
+      "loss": 1.3384,
+      "step": 24504
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.045639753341675,
+      "learning_rate": 4.352498017446471e-05,
+      "loss": 1.3312,
+      "step": 24505
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.712387204170227,
+      "learning_rate": 4.3524715833994184e-05,
+      "loss": 1.2229,
+      "step": 24506
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.4032392501831055,
+      "learning_rate": 4.352445149352366e-05,
+      "loss": 1.233,
+      "step": 24507
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.341879367828369,
+      "learning_rate": 4.352418715305314e-05,
+      "loss": 1.3564,
+      "step": 24508
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0567328929901123,
+      "learning_rate": 4.3523922812582605e-05,
+      "loss": 1.2564,
+      "step": 24509
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.570588231086731,
+      "learning_rate": 4.3523658472112085e-05,
+      "loss": 1.1193,
+      "step": 24510
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.4886211156845093,
+      "learning_rate": 4.352339413164155e-05,
+      "loss": 1.2544,
+      "step": 24511
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0343384742736816,
+      "learning_rate": 4.352312979117103e-05,
+      "loss": 1.3052,
+      "step": 24512
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6927672624588013,
+      "learning_rate": 4.3522865450700506e-05,
+      "loss": 2.1253,
+      "step": 24513
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9344428777694702,
+      "learning_rate": 4.352260111022998e-05,
+      "loss": 1.1543,
+      "step": 24514
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.905179500579834,
+      "learning_rate": 4.352233676975945e-05,
+      "loss": 1.2891,
+      "step": 24515
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.6992244720458984,
+      "learning_rate": 4.3522072429288927e-05,
+      "loss": 1.5219,
+      "step": 24516
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9600411653518677,
+      "learning_rate": 4.35218080888184e-05,
+      "loss": 1.1817,
+      "step": 24517
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.858471155166626,
+      "learning_rate": 4.3521543748347874e-05,
+      "loss": 1.5053,
+      "step": 24518
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.539623975753784,
+      "learning_rate": 4.352127940787735e-05,
+      "loss": 1.3752,
+      "step": 24519
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.120516538619995,
+      "learning_rate": 4.352101506740682e-05,
+      "loss": 1.2605,
+      "step": 24520
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.003999948501587,
+      "learning_rate": 4.3520750726936294e-05,
+      "loss": 1.1549,
+      "step": 24521
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.537167549133301,
+      "learning_rate": 4.352048638646577e-05,
+      "loss": 1.3015,
+      "step": 24522
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.749313235282898,
+      "learning_rate": 4.352022204599525e-05,
+      "loss": 1.1083,
+      "step": 24523
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.5137135982513428,
+      "learning_rate": 4.3519957705524715e-05,
+      "loss": 1.3962,
+      "step": 24524
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.5327706336975098,
+      "learning_rate": 4.3519693365054196e-05,
+      "loss": 1.2457,
+      "step": 24525
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1388027667999268,
+      "learning_rate": 4.351942902458366e-05,
+      "loss": 1.2429,
+      "step": 24526
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.763338565826416,
+      "learning_rate": 4.351916468411314e-05,
+      "loss": 1.3287,
+      "step": 24527
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.7581684589385986,
+      "learning_rate": 4.3518900343642616e-05,
+      "loss": 1.2912,
+      "step": 24528
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6397168636322021,
+      "learning_rate": 4.351863600317208e-05,
+      "loss": 1.222,
+      "step": 24529
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.618316411972046,
+      "learning_rate": 4.351837166270156e-05,
+      "loss": 1.3889,
+      "step": 24530
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1390650272369385,
+      "learning_rate": 4.351810732223103e-05,
+      "loss": 1.1607,
+      "step": 24531
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.9251842498779297,
+      "learning_rate": 4.351784298176051e-05,
+      "loss": 1.3527,
+      "step": 24532
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9406533241271973,
+      "learning_rate": 4.3517578641289984e-05,
+      "loss": 1.1475,
+      "step": 24533
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0758817195892334,
+      "learning_rate": 4.351731430081946e-05,
+      "loss": 1.1402,
+      "step": 24534
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.5545783042907715,
+      "learning_rate": 4.351704996034893e-05,
+      "loss": 1.4017,
+      "step": 24535
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.0216965675354004,
+      "learning_rate": 4.3516785619878405e-05,
+      "loss": 1.4791,
+      "step": 24536
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.169610023498535,
+      "learning_rate": 4.351652127940788e-05,
+      "loss": 1.2945,
+      "step": 24537
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.4014639854431152,
+      "learning_rate": 4.351625693893736e-05,
+      "loss": 1.3509,
+      "step": 24538
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0569558143615723,
+      "learning_rate": 4.3515992598466825e-05,
+      "loss": 1.3322,
+      "step": 24539
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.3967981338500977,
+      "learning_rate": 4.35157282579963e-05,
+      "loss": 1.146,
+      "step": 24540
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.6220057010650635,
+      "learning_rate": 4.351546391752577e-05,
+      "loss": 1.4655,
+      "step": 24541
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.134006977081299,
+      "learning_rate": 4.3515199577055246e-05,
+      "loss": 2.2527,
+      "step": 24542
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9770816564559937,
+      "learning_rate": 4.3514935236584727e-05,
+      "loss": 1.1776,
+      "step": 24543
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6799639463424683,
+      "learning_rate": 4.351467089611419e-05,
+      "loss": 1.2423,
+      "step": 24544
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.8788950443267822,
+      "learning_rate": 4.3514406555643674e-05,
+      "loss": 1.2843,
+      "step": 24545
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1815054416656494,
+      "learning_rate": 4.351414221517314e-05,
+      "loss": 1.3646,
+      "step": 24546
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8533350229263306,
+      "learning_rate": 4.351387787470262e-05,
+      "loss": 1.3527,
+      "step": 24547
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.060009717941284,
+      "learning_rate": 4.3513613534232094e-05,
+      "loss": 1.2504,
+      "step": 24548
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.5138566493988037,
+      "learning_rate": 4.351334919376157e-05,
+      "loss": 1.5271,
+      "step": 24549
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9943211078643799,
+      "learning_rate": 4.351308485329104e-05,
+      "loss": 1.435,
+      "step": 24550
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6898374557495117,
+      "learning_rate": 4.3512820512820515e-05,
+      "loss": 1.224,
+      "step": 24551
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0696375370025635,
+      "learning_rate": 4.351255617234999e-05,
+      "loss": 1.3574,
+      "step": 24552
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9538110494613647,
+      "learning_rate": 4.351229183187946e-05,
+      "loss": 1.3026,
+      "step": 24553
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.2689783573150635,
+      "learning_rate": 4.3512027491408936e-05,
+      "loss": 1.3582,
+      "step": 24554
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.090237855911255,
+      "learning_rate": 4.351176315093841e-05,
+      "loss": 1.2996,
+      "step": 24555
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6491068601608276,
+      "learning_rate": 4.351149881046788e-05,
+      "loss": 1.2681,
+      "step": 24556
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6797116994857788,
+      "learning_rate": 4.3511234469997357e-05,
+      "loss": 1.1465,
+      "step": 24557
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.2101120948791504,
+      "learning_rate": 4.351097012952684e-05,
+      "loss": 1.4759,
+      "step": 24558
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.4816508293151855,
+      "learning_rate": 4.3510705789056304e-05,
+      "loss": 1.3781,
+      "step": 24559
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.300468921661377,
+      "learning_rate": 4.3510441448585784e-05,
+      "loss": 1.3239,
+      "step": 24560
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9336110353469849,
+      "learning_rate": 4.351017710811525e-05,
+      "loss": 1.3834,
+      "step": 24561
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.33573055267334,
+      "learning_rate": 4.350991276764473e-05,
+      "loss": 1.3264,
+      "step": 24562
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.4492039680480957,
+      "learning_rate": 4.3509648427174205e-05,
+      "loss": 1.4324,
+      "step": 24563
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8357945680618286,
+      "learning_rate": 4.350938408670368e-05,
+      "loss": 1.3532,
+      "step": 24564
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.3921167850494385,
+      "learning_rate": 4.350911974623315e-05,
+      "loss": 1.4106,
+      "step": 24565
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.208530902862549,
+      "learning_rate": 4.350885540576262e-05,
+      "loss": 1.3678,
+      "step": 24566
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.2942891120910645,
+      "learning_rate": 4.35085910652921e-05,
+      "loss": 1.2561,
+      "step": 24567
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.769554376602173,
+      "learning_rate": 4.350832672482157e-05,
+      "loss": 1.2828,
+      "step": 24568
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0666704177856445,
+      "learning_rate": 4.3508062384351046e-05,
+      "loss": 1.2218,
+      "step": 24569
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.253716230392456,
+      "learning_rate": 4.350779804388052e-05,
+      "loss": 1.3003,
+      "step": 24570
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.2913758754730225,
+      "learning_rate": 4.350753370340999e-05,
+      "loss": 1.5373,
+      "step": 24571
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.3057456016540527,
+      "learning_rate": 4.350726936293947e-05,
+      "loss": 1.243,
+      "step": 24572
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.310349464416504,
+      "learning_rate": 4.350700502246895e-05,
+      "loss": 2.2843,
+      "step": 24573
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.4317262172698975,
+      "learning_rate": 4.3506740681998414e-05,
+      "loss": 1.5541,
+      "step": 24574
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.091395616531372,
+      "learning_rate": 4.350647634152789e-05,
+      "loss": 1.3834,
+      "step": 24575
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.146754741668701,
+      "learning_rate": 4.350621200105736e-05,
+      "loss": 1.4299,
+      "step": 24576
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8876004219055176,
+      "learning_rate": 4.3505947660586835e-05,
+      "loss": 1.2579,
+      "step": 24577
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.801673650741577,
+      "learning_rate": 4.3505683320116315e-05,
+      "loss": 1.611,
+      "step": 24578
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6915531158447266,
+      "learning_rate": 4.350541897964578e-05,
+      "loss": 1.241,
+      "step": 24579
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.7236509323120117,
+      "learning_rate": 4.350515463917526e-05,
+      "loss": 1.2104,
+      "step": 24580
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0421154499053955,
+      "learning_rate": 4.350489029870473e-05,
+      "loss": 1.4271,
+      "step": 24581
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.339031934738159,
+      "learning_rate": 4.350462595823421e-05,
+      "loss": 1.2614,
+      "step": 24582
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.936144471168518,
+      "learning_rate": 4.350436161776368e-05,
+      "loss": 1.3346,
+      "step": 24583
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.7963252067565918,
+      "learning_rate": 4.3504097277293157e-05,
+      "loss": 1.1502,
+      "step": 24584
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.11854887008667,
+      "learning_rate": 4.350383293682263e-05,
+      "loss": 1.2478,
+      "step": 24585
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9647388458251953,
+      "learning_rate": 4.3503568596352104e-05,
+      "loss": 1.257,
+      "step": 24586
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9925733804702759,
+      "learning_rate": 4.350330425588158e-05,
+      "loss": 1.1723,
+      "step": 24587
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.123340129852295,
+      "learning_rate": 4.350303991541105e-05,
+      "loss": 1.3354,
+      "step": 24588
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6873594522476196,
+      "learning_rate": 4.3502775574940524e-05,
+      "loss": 1.0501,
+      "step": 24589
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.006166458129883,
+      "learning_rate": 4.350251123447e-05,
+      "loss": 1.2656,
+      "step": 24590
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.2057197093963623,
+      "learning_rate": 4.350224689399947e-05,
+      "loss": 1.4393,
+      "step": 24591
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1398518085479736,
+      "learning_rate": 4.3501982553528945e-05,
+      "loss": 1.3222,
+      "step": 24592
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.114898204803467,
+      "learning_rate": 4.3501718213058425e-05,
+      "loss": 1.293,
+      "step": 24593
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.920027256011963,
+      "learning_rate": 4.350145387258789e-05,
+      "loss": 1.2677,
+      "step": 24594
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9163575172424316,
+      "learning_rate": 4.350118953211737e-05,
+      "loss": 1.2543,
+      "step": 24595
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.798844337463379,
+      "learning_rate": 4.350092519164684e-05,
+      "loss": 1.0859,
+      "step": 24596
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.2301230430603027,
+      "learning_rate": 4.350066085117632e-05,
+      "loss": 1.2899,
+      "step": 24597
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.665807008743286,
+      "learning_rate": 4.350039651070579e-05,
+      "loss": 1.5419,
+      "step": 24598
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1570956707000732,
+      "learning_rate": 4.350013217023527e-05,
+      "loss": 1.3444,
+      "step": 24599
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.5235000848770142,
+      "learning_rate": 4.349986782976474e-05,
+      "loss": 1.9309,
+      "step": 24600
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.039907693862915,
+      "learning_rate": 4.349960348929421e-05,
+      "loss": 1.4596,
+      "step": 24601
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9407932758331299,
+      "learning_rate": 4.349933914882369e-05,
+      "loss": 1.2159,
+      "step": 24602
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.5404274463653564,
+      "learning_rate": 4.349907480835316e-05,
+      "loss": 1.5339,
+      "step": 24603
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.081136703491211,
+      "learning_rate": 4.3498810467882635e-05,
+      "loss": 1.3408,
+      "step": 24604
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.002864122390747,
+      "learning_rate": 4.349854612741211e-05,
+      "loss": 1.2902,
+      "step": 24605
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8862309455871582,
+      "learning_rate": 4.349828178694158e-05,
+      "loss": 1.2414,
+      "step": 24606
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8636053800582886,
+      "learning_rate": 4.3498017446471055e-05,
+      "loss": 1.2496,
+      "step": 24607
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1043620109558105,
+      "learning_rate": 4.3497753106000536e-05,
+      "loss": 1.1762,
+      "step": 24608
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.21026349067688,
+      "learning_rate": 4.349748876553e-05,
+      "loss": 1.3022,
+      "step": 24609
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.0199103355407715,
+      "learning_rate": 4.349722442505948e-05,
+      "loss": 1.4193,
+      "step": 24610
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.358391284942627,
+      "learning_rate": 4.349696008458895e-05,
+      "loss": 1.4517,
+      "step": 24611
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.031743049621582,
+      "learning_rate": 4.349669574411842e-05,
+      "loss": 1.2927,
+      "step": 24612
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.010906219482422,
+      "learning_rate": 4.3496431403647904e-05,
+      "loss": 1.171,
+      "step": 24613
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9186853170394897,
+      "learning_rate": 4.349616706317737e-05,
+      "loss": 1.2748,
+      "step": 24614
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6877015829086304,
+      "learning_rate": 4.349590272270685e-05,
+      "loss": 1.2235,
+      "step": 24615
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.6317460536956787,
+      "learning_rate": 4.349563838223632e-05,
+      "loss": 1.3113,
+      "step": 24616
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.025862693786621,
+      "learning_rate": 4.34953740417658e-05,
+      "loss": 1.349,
+      "step": 24617
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0706946849823,
+      "learning_rate": 4.349510970129527e-05,
+      "loss": 1.3425,
+      "step": 24618
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.7950996160507202,
+      "learning_rate": 4.3494845360824745e-05,
+      "loss": 1.1355,
+      "step": 24619
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.5576486587524414,
+      "learning_rate": 4.349458102035422e-05,
+      "loss": 1.3095,
+      "step": 24620
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1408960819244385,
+      "learning_rate": 4.349431667988369e-05,
+      "loss": 1.2265,
+      "step": 24621
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.997952938079834,
+      "learning_rate": 4.3494052339413166e-05,
+      "loss": 1.3901,
+      "step": 24622
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.607515811920166,
+      "learning_rate": 4.349378799894264e-05,
+      "loss": 1.7295,
+      "step": 24623
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.243119478225708,
+      "learning_rate": 4.349352365847211e-05,
+      "loss": 1.286,
+      "step": 24624
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9178152084350586,
+      "learning_rate": 4.3493259318001586e-05,
+      "loss": 1.2283,
+      "step": 24625
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.290102005004883,
+      "learning_rate": 4.349299497753106e-05,
+      "loss": 1.4248,
+      "step": 24626
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6499415636062622,
+      "learning_rate": 4.3492730637060534e-05,
+      "loss": 1.0212,
+      "step": 24627
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.8706789016723633,
+      "learning_rate": 4.3492466296590014e-05,
+      "loss": 1.3344,
+      "step": 24628
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.2308547496795654,
+      "learning_rate": 4.349220195611948e-05,
+      "loss": 1.3575,
+      "step": 24629
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.876503348350525,
+      "learning_rate": 4.349193761564896e-05,
+      "loss": 1.1507,
+      "step": 24630
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9224615097045898,
+      "learning_rate": 4.349167327517843e-05,
+      "loss": 1.2889,
+      "step": 24631
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6217894554138184,
+      "learning_rate": 4.349140893470791e-05,
+      "loss": 1.1136,
+      "step": 24632
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.759584426879883,
+      "learning_rate": 4.349114459423738e-05,
+      "loss": 1.4621,
+      "step": 24633
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9686013460159302,
+      "learning_rate": 4.3490880253766855e-05,
+      "loss": 1.3303,
+      "step": 24634
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.8837289810180664,
+      "learning_rate": 4.349061591329633e-05,
+      "loss": 1.4528,
+      "step": 24635
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8660681247711182,
+      "learning_rate": 4.34903515728258e-05,
+      "loss": 1.2617,
+      "step": 24636
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.328127145767212,
+      "learning_rate": 4.3490087232355276e-05,
+      "loss": 2.1596,
+      "step": 24637
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.4244813919067383,
+      "learning_rate": 4.348982289188475e-05,
+      "loss": 1.2641,
+      "step": 24638
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1927785873413086,
+      "learning_rate": 4.348955855141422e-05,
+      "loss": 1.2925,
+      "step": 24639
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.867310643196106,
+      "learning_rate": 4.34892942109437e-05,
+      "loss": 1.3452,
+      "step": 24640
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1061973571777344,
+      "learning_rate": 4.348902987047317e-05,
+      "loss": 1.5972,
+      "step": 24641
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.135427236557007,
+      "learning_rate": 4.3488765530002644e-05,
+      "loss": 1.3267,
+      "step": 24642
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.02608585357666,
+      "learning_rate": 4.3488501189532124e-05,
+      "loss": 1.3983,
+      "step": 24643
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0042831897735596,
+      "learning_rate": 4.348823684906159e-05,
+      "loss": 1.3081,
+      "step": 24644
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.2870116233825684,
+      "learning_rate": 4.348797250859107e-05,
+      "loss": 1.3399,
+      "step": 24645
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9313098192214966,
+      "learning_rate": 4.348770816812054e-05,
+      "loss": 1.294,
+      "step": 24646
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.022200584411621,
+      "learning_rate": 4.348744382765001e-05,
+      "loss": 1.2498,
+      "step": 24647
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9019508361816406,
+      "learning_rate": 4.348717948717949e-05,
+      "loss": 1.2797,
+      "step": 24648
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6817288398742676,
+      "learning_rate": 4.348691514670896e-05,
+      "loss": 1.309,
+      "step": 24649
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0271050930023193,
+      "learning_rate": 4.348665080623844e-05,
+      "loss": 1.3106,
+      "step": 24650
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.003661870956421,
+      "learning_rate": 4.3486386465767906e-05,
+      "loss": 1.3141,
+      "step": 24651
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.7490959167480469,
+      "learning_rate": 4.3486122125297386e-05,
+      "loss": 1.1884,
+      "step": 24652
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.950395941734314,
+      "learning_rate": 4.348585778482686e-05,
+      "loss": 1.3319,
+      "step": 24653
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.995798945426941,
+      "learning_rate": 4.3485593444356334e-05,
+      "loss": 1.2835,
+      "step": 24654
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.145585775375366,
+      "learning_rate": 4.348532910388581e-05,
+      "loss": 1.3314,
+      "step": 24655
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.5838236808776855,
+      "learning_rate": 4.348506476341528e-05,
+      "loss": 1.4372,
+      "step": 24656
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1595964431762695,
+      "learning_rate": 4.3484800422944754e-05,
+      "loss": 1.4202,
+      "step": 24657
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.943556547164917,
+      "learning_rate": 4.348453608247423e-05,
+      "loss": 1.307,
+      "step": 24658
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.02630877494812,
+      "learning_rate": 4.34842717420037e-05,
+      "loss": 1.2584,
+      "step": 24659
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6359330415725708,
+      "learning_rate": 4.3484007401533175e-05,
+      "loss": 1.2116,
+      "step": 24660
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.7895653247833252,
+      "learning_rate": 4.348374306106265e-05,
+      "loss": 1.3226,
+      "step": 24661
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9150391817092896,
+      "learning_rate": 4.348347872059212e-05,
+      "loss": 1.1121,
+      "step": 24662
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8228119611740112,
+      "learning_rate": 4.34832143801216e-05,
+      "loss": 1.2673,
+      "step": 24663
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8692599534988403,
+      "learning_rate": 4.348295003965107e-05,
+      "loss": 1.2459,
+      "step": 24664
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.019362449645996,
+      "learning_rate": 4.348268569918055e-05,
+      "loss": 1.2927,
+      "step": 24665
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9895516633987427,
+      "learning_rate": 4.3482421358710016e-05,
+      "loss": 1.2697,
+      "step": 24666
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9387537240982056,
+      "learning_rate": 4.34821570182395e-05,
+      "loss": 1.2757,
+      "step": 24667
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8891382217407227,
+      "learning_rate": 4.348189267776897e-05,
+      "loss": 1.2133,
+      "step": 24668
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0703163146972656,
+      "learning_rate": 4.3481628337298444e-05,
+      "loss": 1.2564,
+      "step": 24669
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.646169900894165,
+      "learning_rate": 4.348136399682792e-05,
+      "loss": 1.0948,
+      "step": 24670
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.2023205757141113,
+      "learning_rate": 4.348109965635739e-05,
+      "loss": 1.3484,
+      "step": 24671
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.2689740657806396,
+      "learning_rate": 4.3480835315886865e-05,
+      "loss": 1.3112,
+      "step": 24672
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.678697109222412,
+      "learning_rate": 4.348057097541634e-05,
+      "loss": 1.2726,
+      "step": 24673
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.4836151599884033,
+      "learning_rate": 4.348030663494581e-05,
+      "loss": 1.4265,
+      "step": 24674
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.7620975971221924,
+      "learning_rate": 4.3480042294475285e-05,
+      "loss": 1.3594,
+      "step": 24675
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.878772497177124,
+      "learning_rate": 4.347977795400476e-05,
+      "loss": 1.2116,
+      "step": 24676
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8071171045303345,
+      "learning_rate": 4.347951361353423e-05,
+      "loss": 1.2848,
+      "step": 24677
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.3067474365234375,
+      "learning_rate": 4.347924927306371e-05,
+      "loss": 1.3426,
+      "step": 24678
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9874918460845947,
+      "learning_rate": 4.347898493259318e-05,
+      "loss": 1.3175,
+      "step": 24679
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.912203073501587,
+      "learning_rate": 4.347872059212266e-05,
+      "loss": 1.342,
+      "step": 24680
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6503382921218872,
+      "learning_rate": 4.347845625165213e-05,
+      "loss": 1.2903,
+      "step": 24681
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.177722215652466,
+      "learning_rate": 4.347819191118161e-05,
+      "loss": 1.3318,
+      "step": 24682
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.432122230529785,
+      "learning_rate": 4.347792757071108e-05,
+      "loss": 1.3878,
+      "step": 24683
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.43914532661438,
+      "learning_rate": 4.347766323024055e-05,
+      "loss": 1.2896,
+      "step": 24684
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.433549165725708,
+      "learning_rate": 4.347739888977003e-05,
+      "loss": 1.3447,
+      "step": 24685
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.154611825942993,
+      "learning_rate": 4.3477134549299495e-05,
+      "loss": 1.2638,
+      "step": 24686
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.073469400405884,
+      "learning_rate": 4.3476870208828975e-05,
+      "loss": 1.2392,
+      "step": 24687
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9368979930877686,
+      "learning_rate": 4.347660586835845e-05,
+      "loss": 1.2915,
+      "step": 24688
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8550984859466553,
+      "learning_rate": 4.347634152788792e-05,
+      "loss": 1.2497,
+      "step": 24689
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.322392225265503,
+      "learning_rate": 4.3476077187417396e-05,
+      "loss": 1.2484,
+      "step": 24690
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0604350566864014,
+      "learning_rate": 4.347581284694687e-05,
+      "loss": 1.2875,
+      "step": 24691
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.070112943649292,
+      "learning_rate": 4.347554850647634e-05,
+      "loss": 1.2806,
+      "step": 24692
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.310297727584839,
+      "learning_rate": 4.3475284166005816e-05,
+      "loss": 1.6332,
+      "step": 24693
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.056518316268921,
+      "learning_rate": 4.347501982553529e-05,
+      "loss": 1.2389,
+      "step": 24694
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3591208457946777,
+      "learning_rate": 4.3474755485064764e-05,
+      "loss": 1.293,
+      "step": 24695
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9825778007507324,
+      "learning_rate": 4.347449114459424e-05,
+      "loss": 1.2873,
+      "step": 24696
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.8096158504486084,
+      "learning_rate": 4.347422680412371e-05,
+      "loss": 1.3174,
+      "step": 24697
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7258286476135254,
+      "learning_rate": 4.347396246365319e-05,
+      "loss": 1.2399,
+      "step": 24698
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8118669986724854,
+      "learning_rate": 4.347369812318266e-05,
+      "loss": 1.2809,
+      "step": 24699
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.8874592781066895,
+      "learning_rate": 4.347343378271214e-05,
+      "loss": 1.4383,
+      "step": 24700
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1706976890563965,
+      "learning_rate": 4.3473169442241605e-05,
+      "loss": 1.3138,
+      "step": 24701
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.4858386516571045,
+      "learning_rate": 4.3472905101771085e-05,
+      "loss": 1.4297,
+      "step": 24702
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9917073249816895,
+      "learning_rate": 4.347264076130056e-05,
+      "loss": 1.2952,
+      "step": 24703
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0011370182037354,
+      "learning_rate": 4.347237642083003e-05,
+      "loss": 1.4295,
+      "step": 24704
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.8868422508239746,
+      "learning_rate": 4.3472112080359506e-05,
+      "loss": 1.4136,
+      "step": 24705
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.8037190437316895,
+      "learning_rate": 4.347184773988898e-05,
+      "loss": 1.4058,
+      "step": 24706
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.432001829147339,
+      "learning_rate": 4.347158339941845e-05,
+      "loss": 1.3556,
+      "step": 24707
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9851709604263306,
+      "learning_rate": 4.347131905894793e-05,
+      "loss": 1.2842,
+      "step": 24708
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.147913932800293,
+      "learning_rate": 4.34710547184774e-05,
+      "loss": 1.2783,
+      "step": 24709
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2869324684143066,
+      "learning_rate": 4.3470790378006874e-05,
+      "loss": 1.4406,
+      "step": 24710
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1098005771636963,
+      "learning_rate": 4.347052603753635e-05,
+      "loss": 1.3556,
+      "step": 24711
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.926794409751892,
+      "learning_rate": 4.347026169706582e-05,
+      "loss": 1.8864,
+      "step": 24712
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.108530282974243,
+      "learning_rate": 4.34699973565953e-05,
+      "loss": 1.4069,
+      "step": 24713
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8594882488250732,
+      "learning_rate": 4.346973301612477e-05,
+      "loss": 1.3457,
+      "step": 24714
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.800372838973999,
+      "learning_rate": 4.346946867565425e-05,
+      "loss": 1.2145,
+      "step": 24715
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7234305143356323,
+      "learning_rate": 4.3469204335183715e-05,
+      "loss": 1.2877,
+      "step": 24716
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.122847080230713,
+      "learning_rate": 4.3468939994713196e-05,
+      "loss": 1.4736,
+      "step": 24717
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.819191575050354,
+      "learning_rate": 4.346867565424267e-05,
+      "loss": 1.2681,
+      "step": 24718
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0263025760650635,
+      "learning_rate": 4.3468411313772136e-05,
+      "loss": 1.4229,
+      "step": 24719
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.019312620162964,
+      "learning_rate": 4.3468146973301616e-05,
+      "loss": 1.1832,
+      "step": 24720
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.944035291671753,
+      "learning_rate": 4.346788263283108e-05,
+      "loss": 1.2888,
+      "step": 24721
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2703020572662354,
+      "learning_rate": 4.3467618292360564e-05,
+      "loss": 1.3302,
+      "step": 24722
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.140186309814453,
+      "learning_rate": 4.346735395189004e-05,
+      "loss": 2.1211,
+      "step": 24723
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.5362972021102905,
+      "learning_rate": 4.346708961141951e-05,
+      "loss": 1.6998,
+      "step": 24724
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9305464029312134,
+      "learning_rate": 4.3466825270948984e-05,
+      "loss": 2.22,
+      "step": 24725
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.942921757698059,
+      "learning_rate": 4.346656093047846e-05,
+      "loss": 1.2356,
+      "step": 24726
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3157598972320557,
+      "learning_rate": 4.346629659000793e-05,
+      "loss": 1.1784,
+      "step": 24727
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0035924911499023,
+      "learning_rate": 4.346603224953741e-05,
+      "loss": 1.2329,
+      "step": 24728
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.353567600250244,
+      "learning_rate": 4.346576790906688e-05,
+      "loss": 1.3494,
+      "step": 24729
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0480034351348877,
+      "learning_rate": 4.346550356859635e-05,
+      "loss": 1.2882,
+      "step": 24730
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.6243438720703125,
+      "learning_rate": 4.3465239228125826e-05,
+      "loss": 1.3451,
+      "step": 24731
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6092089414596558,
+      "learning_rate": 4.34649748876553e-05,
+      "loss": 2.1524,
+      "step": 24732
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1037676334381104,
+      "learning_rate": 4.346471054718478e-05,
+      "loss": 1.3302,
+      "step": 24733
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6320505142211914,
+      "learning_rate": 4.3464446206714246e-05,
+      "loss": 1.2544,
+      "step": 24734
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1669692993164062,
+      "learning_rate": 4.346418186624373e-05,
+      "loss": 1.2699,
+      "step": 24735
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9446313381195068,
+      "learning_rate": 4.3463917525773194e-05,
+      "loss": 1.4054,
+      "step": 24736
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.4001073837280273,
+      "learning_rate": 4.3463653185302674e-05,
+      "loss": 1.4291,
+      "step": 24737
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1103010177612305,
+      "learning_rate": 4.346338884483215e-05,
+      "loss": 1.4798,
+      "step": 24738
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0463130474090576,
+      "learning_rate": 4.346312450436162e-05,
+      "loss": 1.2523,
+      "step": 24739
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.5026514530181885,
+      "learning_rate": 4.3462860163891095e-05,
+      "loss": 1.3799,
+      "step": 24740
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.567622184753418,
+      "learning_rate": 4.346259582342057e-05,
+      "loss": 1.3204,
+      "step": 24741
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.905731439590454,
+      "learning_rate": 4.346233148295004e-05,
+      "loss": 1.3232,
+      "step": 24742
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6876535415649414,
+      "learning_rate": 4.3462067142479515e-05,
+      "loss": 1.2859,
+      "step": 24743
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8921326398849487,
+      "learning_rate": 4.346180280200899e-05,
+      "loss": 1.4574,
+      "step": 24744
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8692708015441895,
+      "learning_rate": 4.346153846153846e-05,
+      "loss": 1.2781,
+      "step": 24745
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.038625717163086,
+      "learning_rate": 4.3461274121067936e-05,
+      "loss": 1.2658,
+      "step": 24746
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8464933633804321,
+      "learning_rate": 4.346100978059741e-05,
+      "loss": 1.5117,
+      "step": 24747
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.036733627319336,
+      "learning_rate": 4.346074544012689e-05,
+      "loss": 1.2627,
+      "step": 24748
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.685056686401367,
+      "learning_rate": 4.346048109965636e-05,
+      "loss": 1.3545,
+      "step": 24749
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9052585363388062,
+      "learning_rate": 4.346021675918584e-05,
+      "loss": 1.1982,
+      "step": 24750
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2160537242889404,
+      "learning_rate": 4.3459952418715304e-05,
+      "loss": 1.2681,
+      "step": 24751
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0396621227264404,
+      "learning_rate": 4.3459688078244784e-05,
+      "loss": 1.4011,
+      "step": 24752
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3770484924316406,
+      "learning_rate": 4.345942373777426e-05,
+      "loss": 2.5181,
+      "step": 24753
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.569023847579956,
+      "learning_rate": 4.345915939730373e-05,
+      "loss": 1.0401,
+      "step": 24754
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9302318096160889,
+      "learning_rate": 4.3458895056833205e-05,
+      "loss": 1.2614,
+      "step": 24755
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9780354499816895,
+      "learning_rate": 4.345863071636267e-05,
+      "loss": 1.3061,
+      "step": 24756
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9882831573486328,
+      "learning_rate": 4.345836637589215e-05,
+      "loss": 1.2243,
+      "step": 24757
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8134431838989258,
+      "learning_rate": 4.3458102035421626e-05,
+      "loss": 1.3619,
+      "step": 24758
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3560118675231934,
+      "learning_rate": 4.34578376949511e-05,
+      "loss": 1.5702,
+      "step": 24759
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3947675228118896,
+      "learning_rate": 4.345757335448057e-05,
+      "loss": 1.4567,
+      "step": 24760
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1901891231536865,
+      "learning_rate": 4.3457309014010046e-05,
+      "loss": 1.2918,
+      "step": 24761
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0967631340026855,
+      "learning_rate": 4.345704467353952e-05,
+      "loss": 1.3716,
+      "step": 24762
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8090109825134277,
+      "learning_rate": 4.3456780333069e-05,
+      "loss": 1.3992,
+      "step": 24763
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1161417961120605,
+      "learning_rate": 4.345651599259847e-05,
+      "loss": 1.5021,
+      "step": 24764
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1120879650115967,
+      "learning_rate": 4.345625165212794e-05,
+      "loss": 1.4081,
+      "step": 24765
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.2986794710159302,
+      "learning_rate": 4.3455987311657414e-05,
+      "loss": 1.1648,
+      "step": 24766
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9063069820404053,
+      "learning_rate": 4.345572297118689e-05,
+      "loss": 1.2495,
+      "step": 24767
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.4884722232818604,
+      "learning_rate": 4.345545863071637e-05,
+      "loss": 1.4374,
+      "step": 24768
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6585501432418823,
+      "learning_rate": 4.3455194290245835e-05,
+      "loss": 1.1091,
+      "step": 24769
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1510567665100098,
+      "learning_rate": 4.3454929949775315e-05,
+      "loss": 1.2392,
+      "step": 24770
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.6204919815063477,
+      "learning_rate": 4.345466560930478e-05,
+      "loss": 1.5079,
+      "step": 24771
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.869690179824829,
+      "learning_rate": 4.345440126883426e-05,
+      "loss": 1.047,
+      "step": 24772
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.162372350692749,
+      "learning_rate": 4.3454136928363736e-05,
+      "loss": 1.3274,
+      "step": 24773
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.939821720123291,
+      "learning_rate": 4.345387258789321e-05,
+      "loss": 1.3238,
+      "step": 24774
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9165741205215454,
+      "learning_rate": 4.345360824742268e-05,
+      "loss": 1.3948,
+      "step": 24775
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6707812547683716,
+      "learning_rate": 4.345334390695216e-05,
+      "loss": 1.226,
+      "step": 24776
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.081510066986084,
+      "learning_rate": 4.345307956648163e-05,
+      "loss": 1.2726,
+      "step": 24777
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0311009883880615,
+      "learning_rate": 4.3452815226011104e-05,
+      "loss": 1.3339,
+      "step": 24778
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7810864448547363,
+      "learning_rate": 4.345255088554058e-05,
+      "loss": 1.1464,
+      "step": 24779
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8491466045379639,
+      "learning_rate": 4.345228654507005e-05,
+      "loss": 1.3018,
+      "step": 24780
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.8298888206481934,
+      "learning_rate": 4.3452022204599525e-05,
+      "loss": 1.5169,
+      "step": 24781
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9004782438278198,
+      "learning_rate": 4.3451757864129e-05,
+      "loss": 1.2064,
+      "step": 24782
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.273366928100586,
+      "learning_rate": 4.345149352365848e-05,
+      "loss": 1.2664,
+      "step": 24783
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1926019191741943,
+      "learning_rate": 4.3451229183187945e-05,
+      "loss": 1.4477,
+      "step": 24784
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0295510292053223,
+      "learning_rate": 4.3450964842717426e-05,
+      "loss": 1.1964,
+      "step": 24785
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8709146976470947,
+      "learning_rate": 4.345070050224689e-05,
+      "loss": 1.3151,
+      "step": 24786
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.4880027770996094,
+      "learning_rate": 4.345043616177637e-05,
+      "loss": 2.519,
+      "step": 24787
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9872955083847046,
+      "learning_rate": 4.3450171821305846e-05,
+      "loss": 1.304,
+      "step": 24788
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1390388011932373,
+      "learning_rate": 4.344990748083532e-05,
+      "loss": 1.3748,
+      "step": 24789
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.219299554824829,
+      "learning_rate": 4.3449643140364793e-05,
+      "loss": 2.1243,
+      "step": 24790
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3910601139068604,
+      "learning_rate": 4.344937879989427e-05,
+      "loss": 1.3584,
+      "step": 24791
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9536546468734741,
+      "learning_rate": 4.344911445942374e-05,
+      "loss": 1.2727,
+      "step": 24792
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6935038566589355,
+      "learning_rate": 4.3448850118953214e-05,
+      "loss": 1.117,
+      "step": 24793
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3652970790863037,
+      "learning_rate": 4.344858577848269e-05,
+      "loss": 1.2879,
+      "step": 24794
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8782637119293213,
+      "learning_rate": 4.344832143801216e-05,
+      "loss": 1.216,
+      "step": 24795
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1544346809387207,
+      "learning_rate": 4.3448057097541635e-05,
+      "loss": 1.3972,
+      "step": 24796
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.736315369606018,
+      "learning_rate": 4.344779275707111e-05,
+      "loss": 1.1838,
+      "step": 24797
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6309760808944702,
+      "learning_rate": 4.344752841660059e-05,
+      "loss": 3.6545,
+      "step": 24798
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9419772624969482,
+      "learning_rate": 4.3447264076130056e-05,
+      "loss": 1.3666,
+      "step": 24799
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7403825521469116,
+      "learning_rate": 4.3446999735659536e-05,
+      "loss": 1.2377,
+      "step": 24800
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7539479732513428,
+      "learning_rate": 4.3446735395189e-05,
+      "loss": 1.3508,
+      "step": 24801
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.4277873039245605,
+      "learning_rate": 4.3446471054718476e-05,
+      "loss": 1.4705,
+      "step": 24802
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1807785034179688,
+      "learning_rate": 4.344620671424796e-05,
+      "loss": 1.3021,
+      "step": 24803
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9496268033981323,
+      "learning_rate": 4.3445942373777423e-05,
+      "loss": 1.2097,
+      "step": 24804
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.394871950149536,
+      "learning_rate": 4.3445678033306904e-05,
+      "loss": 1.3867,
+      "step": 24805
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7789828777313232,
+      "learning_rate": 4.344541369283637e-05,
+      "loss": 1.1703,
+      "step": 24806
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2845306396484375,
+      "learning_rate": 4.344514935236585e-05,
+      "loss": 1.269,
+      "step": 24807
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.450446844100952,
+      "learning_rate": 4.3444885011895324e-05,
+      "loss": 1.399,
+      "step": 24808
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.054594039916992,
+      "learning_rate": 4.34446206714248e-05,
+      "loss": 1.309,
+      "step": 24809
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.0399086475372314,
+      "learning_rate": 4.344435633095427e-05,
+      "loss": 1.641,
+      "step": 24810
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8441182374954224,
+      "learning_rate": 4.3444091990483745e-05,
+      "loss": 1.274,
+      "step": 24811
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3630993366241455,
+      "learning_rate": 4.344382765001322e-05,
+      "loss": 1.4527,
+      "step": 24812
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9523110389709473,
+      "learning_rate": 4.344356330954269e-05,
+      "loss": 1.2807,
+      "step": 24813
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.078484058380127,
+      "learning_rate": 4.3443298969072166e-05,
+      "loss": 1.3586,
+      "step": 24814
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.466478109359741,
+      "learning_rate": 4.344303462860164e-05,
+      "loss": 1.3152,
+      "step": 24815
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7810423374176025,
+      "learning_rate": 4.344277028813111e-05,
+      "loss": 1.3072,
+      "step": 24816
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3096389770507812,
+      "learning_rate": 4.344250594766059e-05,
+      "loss": 1.2586,
+      "step": 24817
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3936920166015625,
+      "learning_rate": 4.344224160719007e-05,
+      "loss": 1.5166,
+      "step": 24818
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0900955200195312,
+      "learning_rate": 4.3441977266719534e-05,
+      "loss": 1.2487,
+      "step": 24819
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8741981983184814,
+      "learning_rate": 4.3441712926249014e-05,
+      "loss": 1.4407,
+      "step": 24820
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.36053204536438,
+      "learning_rate": 4.344144858577848e-05,
+      "loss": 1.3496,
+      "step": 24821
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8440613746643066,
+      "learning_rate": 4.344118424530796e-05,
+      "loss": 1.2553,
+      "step": 24822
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1295435428619385,
+      "learning_rate": 4.3440919904837435e-05,
+      "loss": 1.3251,
+      "step": 24823
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.352975845336914,
+      "learning_rate": 4.344065556436691e-05,
+      "loss": 1.4858,
+      "step": 24824
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6196588277816772,
+      "learning_rate": 4.344039122389638e-05,
+      "loss": 1.27,
+      "step": 24825
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.827385187149048,
+      "learning_rate": 4.3440126883425856e-05,
+      "loss": 1.3944,
+      "step": 24826
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.365271806716919,
+      "learning_rate": 4.343986254295533e-05,
+      "loss": 1.5943,
+      "step": 24827
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7483365535736084,
+      "learning_rate": 4.34395982024848e-05,
+      "loss": 1.233,
+      "step": 24828
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8013160228729248,
+      "learning_rate": 4.3439333862014276e-05,
+      "loss": 1.3247,
+      "step": 24829
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.017592430114746,
+      "learning_rate": 4.343906952154375e-05,
+      "loss": 1.3973,
+      "step": 24830
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.029559850692749,
+      "learning_rate": 4.3438805181073223e-05,
+      "loss": 1.341,
+      "step": 24831
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8572973012924194,
+      "learning_rate": 4.34385408406027e-05,
+      "loss": 1.2718,
+      "step": 24832
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.804100513458252,
+      "learning_rate": 4.343827650013218e-05,
+      "loss": 1.2232,
+      "step": 24833
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2159292697906494,
+      "learning_rate": 4.3438012159661644e-05,
+      "loss": 1.3991,
+      "step": 24834
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2064802646636963,
+      "learning_rate": 4.3437747819191124e-05,
+      "loss": 1.2776,
+      "step": 24835
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.4313976764678955,
+      "learning_rate": 4.343748347872059e-05,
+      "loss": 1.4158,
+      "step": 24836
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.774334669113159,
+      "learning_rate": 4.3437219138250065e-05,
+      "loss": 1.5961,
+      "step": 24837
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6417101621627808,
+      "learning_rate": 4.3436954797779545e-05,
+      "loss": 1.2043,
+      "step": 24838
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.819718599319458,
+      "learning_rate": 4.343669045730901e-05,
+      "loss": 1.3526,
+      "step": 24839
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.634737491607666,
+      "learning_rate": 4.343642611683849e-05,
+      "loss": 1.1715,
+      "step": 24840
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9576116800308228,
+      "learning_rate": 4.343616177636796e-05,
+      "loss": 1.2367,
+      "step": 24841
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.4527580738067627,
+      "learning_rate": 4.343589743589744e-05,
+      "loss": 1.4131,
+      "step": 24842
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.4547507762908936,
+      "learning_rate": 4.343563309542691e-05,
+      "loss": 1.5025,
+      "step": 24843
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9095484018325806,
+      "learning_rate": 4.3435368754956387e-05,
+      "loss": 1.2632,
+      "step": 24844
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.5931396484375,
+      "learning_rate": 4.343510441448586e-05,
+      "loss": 1.2885,
+      "step": 24845
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.686256170272827,
+      "learning_rate": 4.3434840074015334e-05,
+      "loss": 1.3886,
+      "step": 24846
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.650948166847229,
+      "learning_rate": 4.343457573354481e-05,
+      "loss": 1.2799,
+      "step": 24847
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9515397548675537,
+      "learning_rate": 4.343431139307428e-05,
+      "loss": 1.4827,
+      "step": 24848
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.984252691268921,
+      "learning_rate": 4.3434047052603754e-05,
+      "loss": 1.2996,
+      "step": 24849
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3116323947906494,
+      "learning_rate": 4.343378271213323e-05,
+      "loss": 1.3701,
+      "step": 24850
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8201571702957153,
+      "learning_rate": 4.34335183716627e-05,
+      "loss": 1.3084,
+      "step": 24851
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.9716603755950928,
+      "learning_rate": 4.3433254031192175e-05,
+      "loss": 1.4667,
+      "step": 24852
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.378267765045166,
+      "learning_rate": 4.3432989690721656e-05,
+      "loss": 1.4072,
+      "step": 24853
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.070328712463379,
+      "learning_rate": 4.343272535025112e-05,
+      "loss": 1.321,
+      "step": 24854
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3179702758789062,
+      "learning_rate": 4.34324610097806e-05,
+      "loss": 1.3225,
+      "step": 24855
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9554342031478882,
+      "learning_rate": 4.343219666931007e-05,
+      "loss": 1.3093,
+      "step": 24856
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.459949493408203,
+      "learning_rate": 4.343193232883955e-05,
+      "loss": 1.5364,
+      "step": 24857
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0431017875671387,
+      "learning_rate": 4.343166798836902e-05,
+      "loss": 1.3327,
+      "step": 24858
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.929553508758545,
+      "learning_rate": 4.34314036478985e-05,
+      "loss": 1.4438,
+      "step": 24859
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.97989022731781,
+      "learning_rate": 4.343113930742797e-05,
+      "loss": 1.3028,
+      "step": 24860
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.018021583557129,
+      "learning_rate": 4.3430874966957444e-05,
+      "loss": 1.32,
+      "step": 24861
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1243815422058105,
+      "learning_rate": 4.343061062648692e-05,
+      "loss": 1.2586,
+      "step": 24862
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.121680498123169,
+      "learning_rate": 4.343034628601639e-05,
+      "loss": 1.3568,
+      "step": 24863
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.038217067718506,
+      "learning_rate": 4.3430081945545865e-05,
+      "loss": 1.2303,
+      "step": 24864
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.8473154306411743,
+      "learning_rate": 4.342981760507534e-05,
+      "loss": 1.3799,
+      "step": 24865
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0578503608703613,
+      "learning_rate": 4.342955326460481e-05,
+      "loss": 1.6628,
+      "step": 24866
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.389292001724243,
+      "learning_rate": 4.3429288924134286e-05,
+      "loss": 1.3236,
+      "step": 24867
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3273723125457764,
+      "learning_rate": 4.3429024583663766e-05,
+      "loss": 1.3365,
+      "step": 24868
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.4834964275360107,
+      "learning_rate": 4.342876024319323e-05,
+      "loss": 1.7635,
+      "step": 24869
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.035060167312622,
+      "learning_rate": 4.342849590272271e-05,
+      "loss": 1.3418,
+      "step": 24870
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.909515142440796,
+      "learning_rate": 4.342823156225218e-05,
+      "loss": 1.3736,
+      "step": 24871
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.510286331176758,
+      "learning_rate": 4.342796722178166e-05,
+      "loss": 1.3001,
+      "step": 24872
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6076732873916626,
+      "learning_rate": 4.3427702881311134e-05,
+      "loss": 1.1517,
+      "step": 24873
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.370741128921509,
+      "learning_rate": 4.34274385408406e-05,
+      "loss": 1.4271,
+      "step": 24874
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.0599801540374756,
+      "learning_rate": 4.342717420037008e-05,
+      "loss": 1.4145,
+      "step": 24875
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.351569175720215,
+      "learning_rate": 4.342690985989955e-05,
+      "loss": 1.3981,
+      "step": 24876
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7718487977981567,
+      "learning_rate": 4.342664551942903e-05,
+      "loss": 1.1014,
+      "step": 24877
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0452818870544434,
+      "learning_rate": 4.34263811789585e-05,
+      "loss": 1.2851,
+      "step": 24878
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7086488008499146,
+      "learning_rate": 4.3426116838487975e-05,
+      "loss": 1.1732,
+      "step": 24879
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9450680017471313,
+      "learning_rate": 4.342585249801745e-05,
+      "loss": 1.3908,
+      "step": 24880
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9742844104766846,
+      "learning_rate": 4.342558815754692e-05,
+      "loss": 1.2412,
+      "step": 24881
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8593254089355469,
+      "learning_rate": 4.3425323817076396e-05,
+      "loss": 1.2176,
+      "step": 24882
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.6109108924865723,
+      "learning_rate": 4.342505947660587e-05,
+      "loss": 1.3972,
+      "step": 24883
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.5915467739105225,
+      "learning_rate": 4.342479513613534e-05,
+      "loss": 1.3225,
+      "step": 24884
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.042630195617676,
+      "learning_rate": 4.3424530795664817e-05,
+      "loss": 1.5539,
+      "step": 24885
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1502089500427246,
+      "learning_rate": 4.342426645519429e-05,
+      "loss": 1.326,
+      "step": 24886
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.82594895362854,
+      "learning_rate": 4.3424002114723764e-05,
+      "loss": 1.3923,
+      "step": 24887
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.017773389816284,
+      "learning_rate": 4.3423737774253244e-05,
+      "loss": 1.4221,
+      "step": 24888
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.931678295135498,
+      "learning_rate": 4.342347343378271e-05,
+      "loss": 1.3885,
+      "step": 24889
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.992875099182129,
+      "learning_rate": 4.342320909331219e-05,
+      "loss": 1.3992,
+      "step": 24890
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.172722816467285,
+      "learning_rate": 4.342294475284166e-05,
+      "loss": 1.3299,
+      "step": 24891
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7659646272659302,
+      "learning_rate": 4.342268041237114e-05,
+      "loss": 1.2147,
+      "step": 24892
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.346923828125,
+      "learning_rate": 4.342241607190061e-05,
+      "loss": 1.3407,
+      "step": 24893
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.5442720651626587,
+      "learning_rate": 4.3422151731430085e-05,
+      "loss": 1.2559,
+      "step": 24894
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.784160852432251,
+      "learning_rate": 4.342188739095956e-05,
+      "loss": 1.449,
+      "step": 24895
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1715030670166016,
+      "learning_rate": 4.342162305048903e-05,
+      "loss": 1.3114,
+      "step": 24896
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.70526123046875,
+      "learning_rate": 4.3421358710018506e-05,
+      "loss": 1.3981,
+      "step": 24897
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1290690898895264,
+      "learning_rate": 4.342109436954798e-05,
+      "loss": 1.2982,
+      "step": 24898
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.8877813816070557,
+      "learning_rate": 4.342083002907745e-05,
+      "loss": 1.2025,
+      "step": 24899
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.141730785369873,
+      "learning_rate": 4.342056568860693e-05,
+      "loss": 1.303,
+      "step": 24900
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.6027626991271973,
+      "learning_rate": 4.34203013481364e-05,
+      "loss": 1.5405,
+      "step": 24901
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.910043716430664,
+      "learning_rate": 4.3420037007665874e-05,
+      "loss": 1.2006,
+      "step": 24902
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.2901768684387207,
+      "learning_rate": 4.3419772667195354e-05,
+      "loss": 1.2669,
+      "step": 24903
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0670089721679688,
+      "learning_rate": 4.341950832672482e-05,
+      "loss": 1.3474,
+      "step": 24904
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.173931360244751,
+      "learning_rate": 4.34192439862543e-05,
+      "loss": 1.1592,
+      "step": 24905
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.3704774379730225,
+      "learning_rate": 4.341897964578377e-05,
+      "loss": 1.3959,
+      "step": 24906
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.011385917663574,
+      "learning_rate": 4.341871530531325e-05,
+      "loss": 2.2528,
+      "step": 24907
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7697792053222656,
+      "learning_rate": 4.341845096484272e-05,
+      "loss": 1.2042,
+      "step": 24908
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8872323036193848,
+      "learning_rate": 4.341818662437219e-05,
+      "loss": 1.287,
+      "step": 24909
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.6467074155807495,
+      "learning_rate": 4.341792228390167e-05,
+      "loss": 1.1309,
+      "step": 24910
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.7372491359710693,
+      "learning_rate": 4.3417657943431136e-05,
+      "loss": 1.561,
+      "step": 24911
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.5227946043014526,
+      "learning_rate": 4.3417393602960617e-05,
+      "loss": 1.1909,
+      "step": 24912
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.005122184753418,
+      "learning_rate": 4.341712926249009e-05,
+      "loss": 1.2208,
+      "step": 24913
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.6197739839553833,
+      "learning_rate": 4.3416864922019564e-05,
+      "loss": 1.2096,
+      "step": 24914
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.34110164642334,
+      "learning_rate": 4.341660058154904e-05,
+      "loss": 1.422,
+      "step": 24915
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0320510864257812,
+      "learning_rate": 4.341633624107851e-05,
+      "loss": 1.4995,
+      "step": 24916
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.6440868377685547,
+      "learning_rate": 4.3416071900607984e-05,
+      "loss": 1.1394,
+      "step": 24917
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.147500514984131,
+      "learning_rate": 4.3415807560137465e-05,
+      "loss": 1.4393,
+      "step": 24918
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9725308418273926,
+      "learning_rate": 4.341554321966693e-05,
+      "loss": 1.2947,
+      "step": 24919
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1354308128356934,
+      "learning_rate": 4.3415278879196405e-05,
+      "loss": 1.2858,
+      "step": 24920
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.4527647495269775,
+      "learning_rate": 4.341501453872588e-05,
+      "loss": 1.3006,
+      "step": 24921
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9770653247833252,
+      "learning_rate": 4.341475019825535e-05,
+      "loss": 1.3509,
+      "step": 24922
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.535216808319092,
+      "learning_rate": 4.341448585778483e-05,
+      "loss": 2.1458,
+      "step": 24923
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.067880153656006,
+      "learning_rate": 4.34142215173143e-05,
+      "loss": 1.4401,
+      "step": 24924
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.755764365196228,
+      "learning_rate": 4.341395717684378e-05,
+      "loss": 1.1929,
+      "step": 24925
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.613440752029419,
+      "learning_rate": 4.3413692836373247e-05,
+      "loss": 1.2875,
+      "step": 24926
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.903441309928894,
+      "learning_rate": 4.341342849590273e-05,
+      "loss": 1.3251,
+      "step": 24927
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.890502691268921,
+      "learning_rate": 4.34131641554322e-05,
+      "loss": 1.2664,
+      "step": 24928
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7987092733383179,
+      "learning_rate": 4.3412899814961674e-05,
+      "loss": 1.3464,
+      "step": 24929
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9792299270629883,
+      "learning_rate": 4.341263547449115e-05,
+      "loss": 1.2601,
+      "step": 24930
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.6970770359039307,
+      "learning_rate": 4.341237113402062e-05,
+      "loss": 1.4637,
+      "step": 24931
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.3565609455108643,
+      "learning_rate": 4.3412106793550095e-05,
+      "loss": 1.3244,
+      "step": 24932
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.581444501876831,
+      "learning_rate": 4.341184245307957e-05,
+      "loss": 1.2703,
+      "step": 24933
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.03302001953125,
+      "learning_rate": 4.341157811260904e-05,
+      "loss": 1.3754,
+      "step": 24934
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.10550856590271,
+      "learning_rate": 4.3411313772138515e-05,
+      "loss": 1.3505,
+      "step": 24935
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.226661443710327,
+      "learning_rate": 4.341104943166799e-05,
+      "loss": 1.3703,
+      "step": 24936
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.4454803466796875,
+      "learning_rate": 4.341078509119746e-05,
+      "loss": 1.4338,
+      "step": 24937
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.259350538253784,
+      "learning_rate": 4.341052075072694e-05,
+      "loss": 1.4486,
+      "step": 24938
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8008089065551758,
+      "learning_rate": 4.341025641025641e-05,
+      "loss": 1.2995,
+      "step": 24939
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1243109703063965,
+      "learning_rate": 4.340999206978589e-05,
+      "loss": 1.3453,
+      "step": 24940
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.684453010559082,
+      "learning_rate": 4.340972772931536e-05,
+      "loss": 1.1332,
+      "step": 24941
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8665969371795654,
+      "learning_rate": 4.340946338884484e-05,
+      "loss": 1.2424,
+      "step": 24942
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1304736137390137,
+      "learning_rate": 4.340919904837431e-05,
+      "loss": 1.2685,
+      "step": 24943
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.573906660079956,
+      "learning_rate": 4.3408934707903784e-05,
+      "loss": 1.0343,
+      "step": 24944
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9103410243988037,
+      "learning_rate": 4.340867036743326e-05,
+      "loss": 1.2979,
+      "step": 24945
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.5689916610717773,
+      "learning_rate": 4.3408406026962725e-05,
+      "loss": 1.1719,
+      "step": 24946
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8480843305587769,
+      "learning_rate": 4.3408141686492205e-05,
+      "loss": 1.2165,
+      "step": 24947
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0753085613250732,
+      "learning_rate": 4.340787734602168e-05,
+      "loss": 1.2702,
+      "step": 24948
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.781477451324463,
+      "learning_rate": 4.340761300555115e-05,
+      "loss": 1.3732,
+      "step": 24949
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.4190759658813477,
+      "learning_rate": 4.3407348665080626e-05,
+      "loss": 1.1106,
+      "step": 24950
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.2742228507995605,
+      "learning_rate": 4.34070843246101e-05,
+      "loss": 1.3455,
+      "step": 24951
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.5086190700531006,
+      "learning_rate": 4.340681998413957e-05,
+      "loss": 1.2659,
+      "step": 24952
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8974288702011108,
+      "learning_rate": 4.340655564366905e-05,
+      "loss": 1.3224,
+      "step": 24953
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.21165132522583,
+      "learning_rate": 4.340629130319852e-05,
+      "loss": 1.3785,
+      "step": 24954
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.4998764991760254,
+      "learning_rate": 4.3406026962727994e-05,
+      "loss": 1.2772,
+      "step": 24955
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7197985649108887,
+      "learning_rate": 4.340576262225747e-05,
+      "loss": 1.2306,
+      "step": 24956
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.3444743156433105,
+      "learning_rate": 4.340549828178694e-05,
+      "loss": 1.2792,
+      "step": 24957
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1534039974212646,
+      "learning_rate": 4.340523394131642e-05,
+      "loss": 1.2051,
+      "step": 24958
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.7455568313598633,
+      "learning_rate": 4.340496960084589e-05,
+      "loss": 1.5147,
+      "step": 24959
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1791679859161377,
+      "learning_rate": 4.340470526037537e-05,
+      "loss": 1.3104,
+      "step": 24960
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.2175326347351074,
+      "learning_rate": 4.3404440919904835e-05,
+      "loss": 1.3807,
+      "step": 24961
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.274538516998291,
+      "learning_rate": 4.3404176579434315e-05,
+      "loss": 1.2992,
+      "step": 24962
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7894787788391113,
+      "learning_rate": 4.340391223896379e-05,
+      "loss": 1.9633,
+      "step": 24963
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.746761679649353,
+      "learning_rate": 4.340364789849326e-05,
+      "loss": 1.7042,
+      "step": 24964
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9762569665908813,
+      "learning_rate": 4.3403383558022736e-05,
+      "loss": 1.2966,
+      "step": 24965
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.214909315109253,
+      "learning_rate": 4.340311921755221e-05,
+      "loss": 1.2493,
+      "step": 24966
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7706035375595093,
+      "learning_rate": 4.340285487708168e-05,
+      "loss": 1.0807,
+      "step": 24967
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8211138248443604,
+      "learning_rate": 4.340259053661116e-05,
+      "loss": 1.2107,
+      "step": 24968
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.6336281299591064,
+      "learning_rate": 4.340232619614063e-05,
+      "loss": 1.1696,
+      "step": 24969
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.113849639892578,
+      "learning_rate": 4.3402061855670104e-05,
+      "loss": 1.3294,
+      "step": 24970
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.289801836013794,
+      "learning_rate": 4.340179751519958e-05,
+      "loss": 1.455,
+      "step": 24971
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.909759759902954,
+      "learning_rate": 4.340153317472905e-05,
+      "loss": 1.4705,
+      "step": 24972
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.512730121612549,
+      "learning_rate": 4.340126883425853e-05,
+      "loss": 1.3839,
+      "step": 24973
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7795900106430054,
+      "learning_rate": 4.3401004493788e-05,
+      "loss": 1.227,
+      "step": 24974
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0135104656219482,
+      "learning_rate": 4.340074015331748e-05,
+      "loss": 1.327,
+      "step": 24975
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.6777782440185547,
+      "learning_rate": 4.3400475812846945e-05,
+      "loss": 1.0955,
+      "step": 24976
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.3927245140075684,
+      "learning_rate": 4.3400211472376426e-05,
+      "loss": 1.3426,
+      "step": 24977
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8241114616394043,
+      "learning_rate": 4.33999471319059e-05,
+      "loss": 1.2873,
+      "step": 24978
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0001533031463623,
+      "learning_rate": 4.339968279143537e-05,
+      "loss": 1.3044,
+      "step": 24979
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.986523985862732,
+      "learning_rate": 4.3399418450964846e-05,
+      "loss": 1.4227,
+      "step": 24980
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.4376180171966553,
+      "learning_rate": 4.339915411049432e-05,
+      "loss": 1.359,
+      "step": 24981
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7009992599487305,
+      "learning_rate": 4.3398889770023794e-05,
+      "loss": 1.2822,
+      "step": 24982
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.4414275884628296,
+      "learning_rate": 4.339862542955327e-05,
+      "loss": 1.0808,
+      "step": 24983
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.841291904449463,
+      "learning_rate": 4.339836108908274e-05,
+      "loss": 1.6358,
+      "step": 24984
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.356207847595215,
+      "learning_rate": 4.3398096748612214e-05,
+      "loss": 1.3323,
+      "step": 24985
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9743331670761108,
+      "learning_rate": 4.339783240814169e-05,
+      "loss": 1.4065,
+      "step": 24986
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7436354160308838,
+      "learning_rate": 4.339756806767116e-05,
+      "loss": 1.2423,
+      "step": 24987
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.285526752471924,
+      "learning_rate": 4.339730372720064e-05,
+      "loss": 1.2216,
+      "step": 24988
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8576416969299316,
+      "learning_rate": 4.339703938673011e-05,
+      "loss": 1.4236,
+      "step": 24989
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.211183547973633,
+      "learning_rate": 4.339677504625959e-05,
+      "loss": 1.3681,
+      "step": 24990
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.291414260864258,
+      "learning_rate": 4.3396510705789056e-05,
+      "loss": 1.425,
+      "step": 24991
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9640371799468994,
+      "learning_rate": 4.339624636531853e-05,
+      "loss": 1.1474,
+      "step": 24992
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.7543392181396484,
+      "learning_rate": 4.339598202484801e-05,
+      "loss": 1.3773,
+      "step": 24993
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.150749444961548,
+      "learning_rate": 4.3395717684377476e-05,
+      "loss": 1.3236,
+      "step": 24994
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.164449691772461,
+      "learning_rate": 4.339545334390696e-05,
+      "loss": 1.4615,
+      "step": 24995
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9385664463043213,
+      "learning_rate": 4.3395189003436424e-05,
+      "loss": 1.2327,
+      "step": 24996
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7622555494308472,
+      "learning_rate": 4.3394924662965904e-05,
+      "loss": 1.1471,
+      "step": 24997
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.086944103240967,
+      "learning_rate": 4.339466032249538e-05,
+      "loss": 1.2958,
+      "step": 24998
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9980690479278564,
+      "learning_rate": 4.339439598202485e-05,
+      "loss": 1.3051,
+      "step": 24999
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.890047311782837,
+      "learning_rate": 4.3394131641554325e-05,
+      "loss": 1.4592,
+      "step": 25000
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7636219263076782,
+      "learning_rate": 4.33938673010838e-05,
+      "loss": 1.3611,
+      "step": 25001
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8837717771530151,
+      "learning_rate": 4.339360296061327e-05,
+      "loss": 1.2797,
+      "step": 25002
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9028278589248657,
+      "learning_rate": 4.3393338620142745e-05,
+      "loss": 1.1443,
+      "step": 25003
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.107408046722412,
+      "learning_rate": 4.339307427967222e-05,
+      "loss": 1.2369,
+      "step": 25004
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.2615902423858643,
+      "learning_rate": 4.339280993920169e-05,
+      "loss": 1.3154,
+      "step": 25005
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7942489385604858,
+      "learning_rate": 4.3392545598731166e-05,
+      "loss": 1.1156,
+      "step": 25006
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.943910837173462,
+      "learning_rate": 4.339228125826064e-05,
+      "loss": 1.3219,
+      "step": 25007
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8640449047088623,
+      "learning_rate": 4.339201691779012e-05,
+      "loss": 1.069,
+      "step": 25008
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9929689168930054,
+      "learning_rate": 4.339175257731959e-05,
+      "loss": 1.4153,
+      "step": 25009
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1745924949645996,
+      "learning_rate": 4.339148823684907e-05,
+      "loss": 1.3339,
+      "step": 25010
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.8749375343322754,
+      "learning_rate": 4.3391223896378534e-05,
+      "loss": 1.4782,
+      "step": 25011
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.36248779296875,
+      "learning_rate": 4.3390959555908014e-05,
+      "loss": 1.2974,
+      "step": 25012
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8430848121643066,
+      "learning_rate": 4.339069521543749e-05,
+      "loss": 1.3212,
+      "step": 25013
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1437554359436035,
+      "learning_rate": 4.339043087496696e-05,
+      "loss": 1.236,
+      "step": 25014
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.942880630493164,
+      "learning_rate": 4.3390166534496435e-05,
+      "loss": 1.275,
+      "step": 25015
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.3869683742523193,
+      "learning_rate": 4.338990219402591e-05,
+      "loss": 1.4653,
+      "step": 25016
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9635827541351318,
+      "learning_rate": 4.338963785355538e-05,
+      "loss": 1.3648,
+      "step": 25017
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.297370672225952,
+      "learning_rate": 4.3389373513084856e-05,
+      "loss": 1.3752,
+      "step": 25018
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9803335666656494,
+      "learning_rate": 4.338910917261433e-05,
+      "loss": 1.3318,
+      "step": 25019
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.234980821609497,
+      "learning_rate": 4.33888448321438e-05,
+      "loss": 1.3672,
+      "step": 25020
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.2582645416259766,
+      "learning_rate": 4.3388580491673276e-05,
+      "loss": 1.1478,
+      "step": 25021
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0245814323425293,
+      "learning_rate": 4.338831615120275e-05,
+      "loss": 1.3355,
+      "step": 25022
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.358264923095703,
+      "learning_rate": 4.338805181073223e-05,
+      "loss": 1.313,
+      "step": 25023
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8104287385940552,
+      "learning_rate": 4.33877874702617e-05,
+      "loss": 1.3031,
+      "step": 25024
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.006718873977661,
+      "learning_rate": 4.338752312979118e-05,
+      "loss": 1.3099,
+      "step": 25025
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7643635272979736,
+      "learning_rate": 4.3387258789320644e-05,
+      "loss": 1.1153,
+      "step": 25026
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.4017698764801025,
+      "learning_rate": 4.338699444885012e-05,
+      "loss": 1.3482,
+      "step": 25027
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.183938980102539,
+      "learning_rate": 4.33867301083796e-05,
+      "loss": 1.3568,
+      "step": 25028
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8914943933486938,
+      "learning_rate": 4.3386465767909065e-05,
+      "loss": 1.2645,
+      "step": 25029
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7719125747680664,
+      "learning_rate": 4.3386201427438545e-05,
+      "loss": 1.1809,
+      "step": 25030
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.20383882522583,
+      "learning_rate": 4.338593708696801e-05,
+      "loss": 1.3322,
+      "step": 25031
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9867303371429443,
+      "learning_rate": 4.338567274649749e-05,
+      "loss": 1.3231,
+      "step": 25032
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.404139280319214,
+      "learning_rate": 4.3385408406026966e-05,
+      "loss": 1.3854,
+      "step": 25033
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0346922874450684,
+      "learning_rate": 4.338514406555644e-05,
+      "loss": 1.4573,
+      "step": 25034
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.1498327255249023,
+      "learning_rate": 4.338487972508591e-05,
+      "loss": 1.3397,
+      "step": 25035
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8693617582321167,
+      "learning_rate": 4.338461538461539e-05,
+      "loss": 1.2996,
+      "step": 25036
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0215396881103516,
+      "learning_rate": 4.338435104414486e-05,
+      "loss": 3.0916,
+      "step": 25037
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.9940037727355957,
+      "learning_rate": 4.3384086703674334e-05,
+      "loss": 1.4356,
+      "step": 25038
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1392946243286133,
+      "learning_rate": 4.338382236320381e-05,
+      "loss": 1.3592,
+      "step": 25039
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.1936593055725098,
+      "learning_rate": 4.338355802273328e-05,
+      "loss": 1.3484,
+      "step": 25040
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.97001051902771,
+      "learning_rate": 4.3383293682262755e-05,
+      "loss": 1.2234,
+      "step": 25041
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7133203744888306,
+      "learning_rate": 4.338302934179223e-05,
+      "loss": 1.1348,
+      "step": 25042
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7171024084091187,
+      "learning_rate": 4.338276500132171e-05,
+      "loss": 1.145,
+      "step": 25043
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7311310768127441,
+      "learning_rate": 4.3382500660851175e-05,
+      "loss": 1.386,
+      "step": 25044
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9622511863708496,
+      "learning_rate": 4.3382236320380656e-05,
+      "loss": 1.2609,
+      "step": 25045
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0776870250701904,
+      "learning_rate": 4.338197197991012e-05,
+      "loss": 1.3006,
+      "step": 25046
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.986118197441101,
+      "learning_rate": 4.33817076394396e-05,
+      "loss": 1.4252,
+      "step": 25047
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.252615451812744,
+      "learning_rate": 4.3381443298969076e-05,
+      "loss": 1.2718,
+      "step": 25048
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.376883029937744,
+      "learning_rate": 4.338117895849855e-05,
+      "loss": 1.5078,
+      "step": 25049
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8012797832489014,
+      "learning_rate": 4.3380914618028024e-05,
+      "loss": 1.2946,
+      "step": 25050
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.574359893798828,
+      "learning_rate": 4.33806502775575e-05,
+      "loss": 1.4237,
+      "step": 25051
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.3254635334014893,
+      "learning_rate": 4.338038593708697e-05,
+      "loss": 1.3838,
+      "step": 25052
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.354275703430176,
+      "learning_rate": 4.3380121596616444e-05,
+      "loss": 1.5926,
+      "step": 25053
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9357550144195557,
+      "learning_rate": 4.337985725614592e-05,
+      "loss": 1.8793,
+      "step": 25054
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.412684440612793,
+      "learning_rate": 4.337959291567539e-05,
+      "loss": 1.4523,
+      "step": 25055
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8721097707748413,
+      "learning_rate": 4.3379328575204865e-05,
+      "loss": 1.1864,
+      "step": 25056
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.403334379196167,
+      "learning_rate": 4.337906423473434e-05,
+      "loss": 1.3786,
+      "step": 25057
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.001708745956421,
+      "learning_rate": 4.337879989426382e-05,
+      "loss": 1.3011,
+      "step": 25058
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.244903326034546,
+      "learning_rate": 4.3378535553793286e-05,
+      "loss": 1.3921,
+      "step": 25059
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8065625429153442,
+      "learning_rate": 4.3378271213322766e-05,
+      "loss": 1.1678,
+      "step": 25060
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8950718641281128,
+      "learning_rate": 4.337800687285223e-05,
+      "loss": 1.2379,
+      "step": 25061
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.8469390869140625,
+      "learning_rate": 4.337774253238171e-05,
+      "loss": 1.3252,
+      "step": 25062
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.6827632188796997,
+      "learning_rate": 4.337747819191119e-05,
+      "loss": 1.3038,
+      "step": 25063
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.2487192153930664,
+      "learning_rate": 4.3377213851440654e-05,
+      "loss": 1.2644,
+      "step": 25064
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.786137580871582,
+      "learning_rate": 4.3376949510970134e-05,
+      "loss": 1.3224,
+      "step": 25065
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.3338706493377686,
+      "learning_rate": 4.33766851704996e-05,
+      "loss": 1.3981,
+      "step": 25066
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7635879516601562,
+      "learning_rate": 4.337642083002908e-05,
+      "loss": 1.2407,
+      "step": 25067
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.5326805114746094,
+      "learning_rate": 4.3376156489558555e-05,
+      "loss": 1.3753,
+      "step": 25068
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.9042460918426514,
+      "learning_rate": 4.337589214908803e-05,
+      "loss": 1.3894,
+      "step": 25069
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.232808828353882,
+      "learning_rate": 4.33756278086175e-05,
+      "loss": 1.777,
+      "step": 25070
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.297525644302368,
+      "learning_rate": 4.3375363468146975e-05,
+      "loss": 1.2665,
+      "step": 25071
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8498480319976807,
+      "learning_rate": 4.337509912767645e-05,
+      "loss": 1.264,
+      "step": 25072
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0634312629699707,
+      "learning_rate": 4.337483478720592e-05,
+      "loss": 1.3379,
+      "step": 25073
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9196168184280396,
+      "learning_rate": 4.3374570446735396e-05,
+      "loss": 1.8546,
+      "step": 25074
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8762919902801514,
+      "learning_rate": 4.337430610626487e-05,
+      "loss": 1.3466,
+      "step": 25075
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0611886978149414,
+      "learning_rate": 4.337404176579434e-05,
+      "loss": 1.3607,
+      "step": 25076
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.4196081161499023,
+      "learning_rate": 4.337377742532382e-05,
+      "loss": 1.3456,
+      "step": 25077
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8525279760360718,
+      "learning_rate": 4.33735130848533e-05,
+      "loss": 1.311,
+      "step": 25078
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.4102261066436768,
+      "learning_rate": 4.3373248744382764e-05,
+      "loss": 1.2618,
+      "step": 25079
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.999768853187561,
+      "learning_rate": 4.3372984403912244e-05,
+      "loss": 1.3887,
+      "step": 25080
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9017298221588135,
+      "learning_rate": 4.337272006344171e-05,
+      "loss": 1.3423,
+      "step": 25081
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0414958000183105,
+      "learning_rate": 4.337245572297119e-05,
+      "loss": 1.3224,
+      "step": 25082
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8625147342681885,
+      "learning_rate": 4.3372191382500665e-05,
+      "loss": 1.3757,
+      "step": 25083
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.2528014183044434,
+      "learning_rate": 4.337192704203014e-05,
+      "loss": 1.5771,
+      "step": 25084
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8501874208450317,
+      "learning_rate": 4.337166270155961e-05,
+      "loss": 1.2095,
+      "step": 25085
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.341818332672119,
+      "learning_rate": 4.3371398361089086e-05,
+      "loss": 1.1707,
+      "step": 25086
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.064070224761963,
+      "learning_rate": 4.337113402061856e-05,
+      "loss": 1.4141,
+      "step": 25087
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.740098476409912,
+      "learning_rate": 4.337086968014803e-05,
+      "loss": 1.3604,
+      "step": 25088
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.418341875076294,
+      "learning_rate": 4.3370605339677506e-05,
+      "loss": 1.3607,
+      "step": 25089
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.172773838043213,
+      "learning_rate": 4.337034099920698e-05,
+      "loss": 1.9612,
+      "step": 25090
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.929220199584961,
+      "learning_rate": 4.3370076658736453e-05,
+      "loss": 1.1563,
+      "step": 25091
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.242218255996704,
+      "learning_rate": 4.336981231826593e-05,
+      "loss": 1.2792,
+      "step": 25092
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9698153734207153,
+      "learning_rate": 4.336954797779541e-05,
+      "loss": 1.3948,
+      "step": 25093
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0310750007629395,
+      "learning_rate": 4.3369283637324874e-05,
+      "loss": 1.4539,
+      "step": 25094
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.138594627380371,
+      "learning_rate": 4.3369019296854355e-05,
+      "loss": 1.3676,
+      "step": 25095
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0760815143585205,
+      "learning_rate": 4.336875495638382e-05,
+      "loss": 1.1938,
+      "step": 25096
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7277709245681763,
+      "learning_rate": 4.33684906159133e-05,
+      "loss": 1.1565,
+      "step": 25097
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0754036903381348,
+      "learning_rate": 4.3368226275442775e-05,
+      "loss": 1.3623,
+      "step": 25098
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.488720178604126,
+      "learning_rate": 4.336796193497225e-05,
+      "loss": 1.4808,
+      "step": 25099
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.212338447570801,
+      "learning_rate": 4.336769759450172e-05,
+      "loss": 1.4216,
+      "step": 25100
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.803439736366272,
+      "learning_rate": 4.336743325403119e-05,
+      "loss": 1.2348,
+      "step": 25101
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6443026065826416,
+      "learning_rate": 4.336716891356067e-05,
+      "loss": 1.2392,
+      "step": 25102
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.940032958984375,
+      "learning_rate": 4.336690457309014e-05,
+      "loss": 1.2236,
+      "step": 25103
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9653770923614502,
+      "learning_rate": 4.336664023261962e-05,
+      "loss": 1.3102,
+      "step": 25104
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.2090749740600586,
+      "learning_rate": 4.336637589214909e-05,
+      "loss": 1.4625,
+      "step": 25105
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.3351151943206787,
+      "learning_rate": 4.3366111551678564e-05,
+      "loss": 1.4022,
+      "step": 25106
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8906328678131104,
+      "learning_rate": 4.336584721120804e-05,
+      "loss": 1.1534,
+      "step": 25107
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.99289870262146,
+      "learning_rate": 4.336558287073752e-05,
+      "loss": 1.2502,
+      "step": 25108
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7066980600357056,
+      "learning_rate": 4.3365318530266985e-05,
+      "loss": 1.0693,
+      "step": 25109
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1050820350646973,
+      "learning_rate": 4.336505418979646e-05,
+      "loss": 1.3091,
+      "step": 25110
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.998178243637085,
+      "learning_rate": 4.336478984932593e-05,
+      "loss": 1.5661,
+      "step": 25111
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.935977578163147,
+      "learning_rate": 4.3364525508855405e-05,
+      "loss": 1.9336,
+      "step": 25112
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9824035167694092,
+      "learning_rate": 4.3364261168384886e-05,
+      "loss": 1.1976,
+      "step": 25113
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.656334400177002,
+      "learning_rate": 4.336399682791435e-05,
+      "loss": 1.307,
+      "step": 25114
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.49527645111084,
+      "learning_rate": 4.336373248744383e-05,
+      "loss": 1.342,
+      "step": 25115
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8212687969207764,
+      "learning_rate": 4.33634681469733e-05,
+      "loss": 1.3246,
+      "step": 25116
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9272408485412598,
+      "learning_rate": 4.336320380650278e-05,
+      "loss": 1.161,
+      "step": 25117
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8067680597305298,
+      "learning_rate": 4.3362939466032253e-05,
+      "loss": 1.198,
+      "step": 25118
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0748167037963867,
+      "learning_rate": 4.336267512556173e-05,
+      "loss": 1.379,
+      "step": 25119
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.287276029586792,
+      "learning_rate": 4.33624107850912e-05,
+      "loss": 1.4252,
+      "step": 25120
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1204628944396973,
+      "learning_rate": 4.3362146444620674e-05,
+      "loss": 1.3133,
+      "step": 25121
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.843554973602295,
+      "learning_rate": 4.336188210415015e-05,
+      "loss": 1.3551,
+      "step": 25122
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.5186030864715576,
+      "learning_rate": 4.336161776367962e-05,
+      "loss": 1.4131,
+      "step": 25123
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.039527654647827,
+      "learning_rate": 4.3361353423209095e-05,
+      "loss": 1.294,
+      "step": 25124
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1825506687164307,
+      "learning_rate": 4.336108908273857e-05,
+      "loss": 1.3637,
+      "step": 25125
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.4257633686065674,
+      "learning_rate": 4.336082474226804e-05,
+      "loss": 1.2887,
+      "step": 25126
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8664312362670898,
+      "learning_rate": 4.3360560401797516e-05,
+      "loss": 1.2631,
+      "step": 25127
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.3706748485565186,
+      "learning_rate": 4.3360296061326996e-05,
+      "loss": 1.3328,
+      "step": 25128
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9065746068954468,
+      "learning_rate": 4.336003172085646e-05,
+      "loss": 1.2469,
+      "step": 25129
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6594069004058838,
+      "learning_rate": 4.335976738038594e-05,
+      "loss": 1.1765,
+      "step": 25130
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6853060722351074,
+      "learning_rate": 4.335950303991541e-05,
+      "loss": 1.1801,
+      "step": 25131
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.905793309211731,
+      "learning_rate": 4.335923869944489e-05,
+      "loss": 1.2818,
+      "step": 25132
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.2959907054901123,
+      "learning_rate": 4.3358974358974364e-05,
+      "loss": 1.2904,
+      "step": 25133
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0438895225524902,
+      "learning_rate": 4.335871001850384e-05,
+      "loss": 1.2744,
+      "step": 25134
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.4763166904449463,
+      "learning_rate": 4.335844567803331e-05,
+      "loss": 1.43,
+      "step": 25135
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7985591888427734,
+      "learning_rate": 4.335818133756278e-05,
+      "loss": 1.2236,
+      "step": 25136
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.3340446949005127,
+      "learning_rate": 4.335791699709226e-05,
+      "loss": 1.2523,
+      "step": 25137
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.600178599357605,
+      "learning_rate": 4.335765265662173e-05,
+      "loss": 1.1203,
+      "step": 25138
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.023036241531372,
+      "learning_rate": 4.3357388316151205e-05,
+      "loss": 1.3944,
+      "step": 25139
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9646352529525757,
+      "learning_rate": 4.335712397568068e-05,
+      "loss": 1.2151,
+      "step": 25140
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.052926778793335,
+      "learning_rate": 4.335685963521015e-05,
+      "loss": 1.4042,
+      "step": 25141
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.551516056060791,
+      "learning_rate": 4.3356595294739626e-05,
+      "loss": 1.6678,
+      "step": 25142
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.3400070667266846,
+      "learning_rate": 4.3356330954269106e-05,
+      "loss": 1.2845,
+      "step": 25143
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.896451473236084,
+      "learning_rate": 4.335606661379857e-05,
+      "loss": 1.227,
+      "step": 25144
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1615099906921387,
+      "learning_rate": 4.335580227332805e-05,
+      "loss": 1.389,
+      "step": 25145
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.19012713432312,
+      "learning_rate": 4.335553793285752e-05,
+      "loss": 1.2664,
+      "step": 25146
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9068583250045776,
+      "learning_rate": 4.3355273592386994e-05,
+      "loss": 1.4431,
+      "step": 25147
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1518495082855225,
+      "learning_rate": 4.3355009251916474e-05,
+      "loss": 1.2178,
+      "step": 25148
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6418166160583496,
+      "learning_rate": 4.335474491144594e-05,
+      "loss": 2.9349,
+      "step": 25149
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.911091923713684,
+      "learning_rate": 4.335448057097542e-05,
+      "loss": 1.3376,
+      "step": 25150
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7559411525726318,
+      "learning_rate": 4.335421623050489e-05,
+      "loss": 1.0472,
+      "step": 25151
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9927122592926025,
+      "learning_rate": 4.335395189003437e-05,
+      "loss": 1.3636,
+      "step": 25152
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1951329708099365,
+      "learning_rate": 4.335368754956384e-05,
+      "loss": 1.3386,
+      "step": 25153
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0581789016723633,
+      "learning_rate": 4.3353423209093316e-05,
+      "loss": 1.3579,
+      "step": 25154
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9366353750228882,
+      "learning_rate": 4.335315886862279e-05,
+      "loss": 1.5189,
+      "step": 25155
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.4339065551757812,
+      "learning_rate": 4.335289452815226e-05,
+      "loss": 1.2384,
+      "step": 25156
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.012956380844116,
+      "learning_rate": 4.3352630187681736e-05,
+      "loss": 1.3765,
+      "step": 25157
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6254162788391113,
+      "learning_rate": 4.335236584721121e-05,
+      "loss": 1.3019,
+      "step": 25158
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9562348127365112,
+      "learning_rate": 4.3352101506740683e-05,
+      "loss": 1.3232,
+      "step": 25159
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.894099473953247,
+      "learning_rate": 4.335183716627016e-05,
+      "loss": 1.306,
+      "step": 25160
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.4692039489746094,
+      "learning_rate": 4.335157282579963e-05,
+      "loss": 1.1362,
+      "step": 25161
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9381846189498901,
+      "learning_rate": 4.3351308485329104e-05,
+      "loss": 2.3979,
+      "step": 25162
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1621251106262207,
+      "learning_rate": 4.3351044144858584e-05,
+      "loss": 1.3671,
+      "step": 25163
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.109433889389038,
+      "learning_rate": 4.335077980438805e-05,
+      "loss": 1.2328,
+      "step": 25164
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.627603530883789,
+      "learning_rate": 4.335051546391753e-05,
+      "loss": 1.4729,
+      "step": 25165
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9932209253311157,
+      "learning_rate": 4.3350251123447e-05,
+      "loss": 1.3846,
+      "step": 25166
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.290410280227661,
+      "learning_rate": 4.334998678297648e-05,
+      "loss": 1.5227,
+      "step": 25167
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.946038842201233,
+      "learning_rate": 4.334972244250595e-05,
+      "loss": 1.3041,
+      "step": 25168
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8413972854614258,
+      "learning_rate": 4.3349458102035426e-05,
+      "loss": 1.2322,
+      "step": 25169
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.067723274230957,
+      "learning_rate": 4.33491937615649e-05,
+      "loss": 1.4719,
+      "step": 25170
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7943638563156128,
+      "learning_rate": 4.334892942109437e-05,
+      "loss": 1.2927,
+      "step": 25171
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.4912493228912354,
+      "learning_rate": 4.334866508062385e-05,
+      "loss": 1.4899,
+      "step": 25172
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7404667139053345,
+      "learning_rate": 4.334840074015332e-05,
+      "loss": 1.2812,
+      "step": 25173
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.03249192237854,
+      "learning_rate": 4.3348136399682794e-05,
+      "loss": 1.225,
+      "step": 25174
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1626691818237305,
+      "learning_rate": 4.334787205921227e-05,
+      "loss": 1.2707,
+      "step": 25175
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.053964614868164,
+      "learning_rate": 4.334760771874174e-05,
+      "loss": 1.2206,
+      "step": 25176
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1170337200164795,
+      "learning_rate": 4.3347343378271214e-05,
+      "loss": 1.1599,
+      "step": 25177
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.770437479019165,
+      "learning_rate": 4.3347079037800695e-05,
+      "loss": 1.2106,
+      "step": 25178
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7094154357910156,
+      "learning_rate": 4.334681469733016e-05,
+      "loss": 1.1473,
+      "step": 25179
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.6475987434387207,
+      "learning_rate": 4.334655035685964e-05,
+      "loss": 2.0505,
+      "step": 25180
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.505681276321411,
+      "learning_rate": 4.334628601638911e-05,
+      "loss": 1.5259,
+      "step": 25181
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.852418065071106,
+      "learning_rate": 4.334602167591858e-05,
+      "loss": 1.2473,
+      "step": 25182
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8654168844223022,
+      "learning_rate": 4.334575733544806e-05,
+      "loss": 1.1971,
+      "step": 25183
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9811586141586304,
+      "learning_rate": 4.334549299497753e-05,
+      "loss": 1.2805,
+      "step": 25184
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.082070827484131,
+      "learning_rate": 4.334522865450701e-05,
+      "loss": 2.3102,
+      "step": 25185
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1097068786621094,
+      "learning_rate": 4.3344964314036477e-05,
+      "loss": 1.404,
+      "step": 25186
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9757952690124512,
+      "learning_rate": 4.334469997356596e-05,
+      "loss": 1.343,
+      "step": 25187
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.2486743927001953,
+      "learning_rate": 4.334443563309543e-05,
+      "loss": 1.2277,
+      "step": 25188
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9862251281738281,
+      "learning_rate": 4.3344171292624904e-05,
+      "loss": 1.3631,
+      "step": 25189
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.98762845993042,
+      "learning_rate": 4.334390695215438e-05,
+      "loss": 1.3744,
+      "step": 25190
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.882577419281006,
+      "learning_rate": 4.334364261168385e-05,
+      "loss": 1.2722,
+      "step": 25191
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.3363215923309326,
+      "learning_rate": 4.3343378271213325e-05,
+      "loss": 1.3286,
+      "step": 25192
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.000544548034668,
+      "learning_rate": 4.33431139307428e-05,
+      "loss": 1.3824,
+      "step": 25193
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.124755382537842,
+      "learning_rate": 4.334284959027227e-05,
+      "loss": 1.3158,
+      "step": 25194
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.722226858139038,
+      "learning_rate": 4.3342585249801746e-05,
+      "loss": 1.3131,
+      "step": 25195
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0774893760681152,
+      "learning_rate": 4.334232090933122e-05,
+      "loss": 1.3039,
+      "step": 25196
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.707327365875244,
+      "learning_rate": 4.334205656886069e-05,
+      "loss": 1.315,
+      "step": 25197
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.014023542404175,
+      "learning_rate": 4.334179222839017e-05,
+      "loss": 1.227,
+      "step": 25198
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1415116786956787,
+      "learning_rate": 4.334152788791964e-05,
+      "loss": 1.2743,
+      "step": 25199
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6765891313552856,
+      "learning_rate": 4.334126354744912e-05,
+      "loss": 1.2208,
+      "step": 25200
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9038480520248413,
+      "learning_rate": 4.334099920697859e-05,
+      "loss": 1.1751,
+      "step": 25201
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.2578375339508057,
+      "learning_rate": 4.334073486650807e-05,
+      "loss": 1.3638,
+      "step": 25202
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.947462320327759,
+      "learning_rate": 4.334047052603754e-05,
+      "loss": 1.3655,
+      "step": 25203
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.805703043937683,
+      "learning_rate": 4.3340206185567014e-05,
+      "loss": 1.2662,
+      "step": 25204
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.8825573921203613,
+      "learning_rate": 4.333994184509649e-05,
+      "loss": 2.0423,
+      "step": 25205
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.9637391567230225,
+      "learning_rate": 4.333967750462596e-05,
+      "loss": 1.2849,
+      "step": 25206
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6678673028945923,
+      "learning_rate": 4.3339413164155435e-05,
+      "loss": 1.2783,
+      "step": 25207
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8976173400878906,
+      "learning_rate": 4.333914882368491e-05,
+      "loss": 1.2075,
+      "step": 25208
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.471951723098755,
+      "learning_rate": 4.333888448321438e-05,
+      "loss": 1.3443,
+      "step": 25209
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9344009160995483,
+      "learning_rate": 4.3338620142743856e-05,
+      "loss": 1.2258,
+      "step": 25210
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.604062557220459,
+      "learning_rate": 4.333835580227333e-05,
+      "loss": 1.2338,
+      "step": 25211
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.094278573989868,
+      "learning_rate": 4.33380914618028e-05,
+      "loss": 1.2875,
+      "step": 25212
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7656413316726685,
+      "learning_rate": 4.333782712133228e-05,
+      "loss": 1.3296,
+      "step": 25213
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.327489137649536,
+      "learning_rate": 4.333756278086175e-05,
+      "loss": 1.3908,
+      "step": 25214
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9472291469573975,
+      "learning_rate": 4.333729844039123e-05,
+      "loss": 1.3099,
+      "step": 25215
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.3031821250915527,
+      "learning_rate": 4.33370340999207e-05,
+      "loss": 1.3866,
+      "step": 25216
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8439321517944336,
+      "learning_rate": 4.333676975945018e-05,
+      "loss": 1.3638,
+      "step": 25217
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7132779359817505,
+      "learning_rate": 4.333650541897965e-05,
+      "loss": 1.3122,
+      "step": 25218
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.6783382892608643,
+      "learning_rate": 4.333624107850912e-05,
+      "loss": 1.303,
+      "step": 25219
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.92816162109375,
+      "learning_rate": 4.33359767380386e-05,
+      "loss": 1.4026,
+      "step": 25220
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.2968337535858154,
+      "learning_rate": 4.3335712397568065e-05,
+      "loss": 1.4257,
+      "step": 25221
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0720465183258057,
+      "learning_rate": 4.3335448057097545e-05,
+      "loss": 1.2436,
+      "step": 25222
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0515570640563965,
+      "learning_rate": 4.333518371662702e-05,
+      "loss": 1.294,
+      "step": 25223
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8503429889678955,
+      "learning_rate": 4.333491937615649e-05,
+      "loss": 1.206,
+      "step": 25224
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7994906902313232,
+      "learning_rate": 4.3334655035685966e-05,
+      "loss": 1.3183,
+      "step": 25225
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8015122413635254,
+      "learning_rate": 4.333439069521544e-05,
+      "loss": 1.2013,
+      "step": 25226
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.984571099281311,
+      "learning_rate": 4.333412635474491e-05,
+      "loss": 1.3676,
+      "step": 25227
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.159759283065796,
+      "learning_rate": 4.333386201427439e-05,
+      "loss": 1.2377,
+      "step": 25228
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6284818649291992,
+      "learning_rate": 4.333359767380386e-05,
+      "loss": 1.1796,
+      "step": 25229
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.2483420372009277,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 1.4048,
+      "step": 25230
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8601268529891968,
+      "learning_rate": 4.333306899286281e-05,
+      "loss": 1.1381,
+      "step": 25231
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8877875804901123,
+      "learning_rate": 4.333280465239228e-05,
+      "loss": 1.3246,
+      "step": 25232
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.085326671600342,
+      "learning_rate": 4.333254031192176e-05,
+      "loss": 1.436,
+      "step": 25233
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.426464557647705,
+      "learning_rate": 4.333227597145123e-05,
+      "loss": 1.3434,
+      "step": 25234
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1008989810943604,
+      "learning_rate": 4.333201163098071e-05,
+      "loss": 1.3514,
+      "step": 25235
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.306173801422119,
+      "learning_rate": 4.3331747290510175e-05,
+      "loss": 1.4359,
+      "step": 25236
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9504050016403198,
+      "learning_rate": 4.3331482950039656e-05,
+      "loss": 1.3271,
+      "step": 25237
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8318732976913452,
+      "learning_rate": 4.333121860956913e-05,
+      "loss": 1.2828,
+      "step": 25238
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8375095129013062,
+      "learning_rate": 4.33309542690986e-05,
+      "loss": 1.1673,
+      "step": 25239
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0624606609344482,
+      "learning_rate": 4.3330689928628077e-05,
+      "loss": 1.2467,
+      "step": 25240
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.3800997734069824,
+      "learning_rate": 4.333042558815755e-05,
+      "loss": 1.3847,
+      "step": 25241
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.8195629119873047,
+      "learning_rate": 4.3330161247687024e-05,
+      "loss": 1.4259,
+      "step": 25242
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.523253917694092,
+      "learning_rate": 4.33298969072165e-05,
+      "loss": 1.4958,
+      "step": 25243
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8014382123947144,
+      "learning_rate": 4.332963256674597e-05,
+      "loss": 2.1457,
+      "step": 25244
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1508381366729736,
+      "learning_rate": 4.3329368226275444e-05,
+      "loss": 1.2745,
+      "step": 25245
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.193488836288452,
+      "learning_rate": 4.332910388580492e-05,
+      "loss": 1.2968,
+      "step": 25246
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.4524106979370117,
+      "learning_rate": 4.332883954533439e-05,
+      "loss": 1.5449,
+      "step": 25247
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1501073837280273,
+      "learning_rate": 4.332857520486387e-05,
+      "loss": 1.2278,
+      "step": 25248
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.4071247577667236,
+      "learning_rate": 4.332831086439334e-05,
+      "loss": 1.3766,
+      "step": 25249
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.724086880683899,
+      "learning_rate": 4.332804652392282e-05,
+      "loss": 1.1426,
+      "step": 25250
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.993986964225769,
+      "learning_rate": 4.3327782183452286e-05,
+      "loss": 1.5771,
+      "step": 25251
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1091792583465576,
+      "learning_rate": 4.3327517842981766e-05,
+      "loss": 1.3835,
+      "step": 25252
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0902321338653564,
+      "learning_rate": 4.332725350251124e-05,
+      "loss": 1.3331,
+      "step": 25253
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9767439365386963,
+      "learning_rate": 4.3326989162040707e-05,
+      "loss": 1.3173,
+      "step": 25254
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2380101680755615,
+      "learning_rate": 4.332672482157019e-05,
+      "loss": 1.3827,
+      "step": 25255
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.0906789302825928,
+      "learning_rate": 4.3326460481099654e-05,
+      "loss": 1.4898,
+      "step": 25256
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7217603921890259,
+      "learning_rate": 4.3326196140629134e-05,
+      "loss": 1.2973,
+      "step": 25257
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9374245405197144,
+      "learning_rate": 4.332593180015861e-05,
+      "loss": 1.2263,
+      "step": 25258
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9026682376861572,
+      "learning_rate": 4.332566745968808e-05,
+      "loss": 1.2561,
+      "step": 25259
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.659722328186035,
+      "learning_rate": 4.3325403119217555e-05,
+      "loss": 1.3735,
+      "step": 25260
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1556899547576904,
+      "learning_rate": 4.332513877874703e-05,
+      "loss": 1.3384,
+      "step": 25261
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.4689126014709473,
+      "learning_rate": 4.33248744382765e-05,
+      "loss": 1.1125,
+      "step": 25262
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3565566539764404,
+      "learning_rate": 4.3324610097805975e-05,
+      "loss": 1.2291,
+      "step": 25263
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8157877922058105,
+      "learning_rate": 4.332434575733545e-05,
+      "loss": 1.3265,
+      "step": 25264
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3124310970306396,
+      "learning_rate": 4.332408141686492e-05,
+      "loss": 1.4687,
+      "step": 25265
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7515596151351929,
+      "learning_rate": 4.3323817076394396e-05,
+      "loss": 1.2466,
+      "step": 25266
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1768927574157715,
+      "learning_rate": 4.332355273592387e-05,
+      "loss": 1.209,
+      "step": 25267
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8830687999725342,
+      "learning_rate": 4.332328839545335e-05,
+      "loss": 1.318,
+      "step": 25268
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7506829500198364,
+      "learning_rate": 4.332302405498282e-05,
+      "loss": 1.1881,
+      "step": 25269
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8754035234451294,
+      "learning_rate": 4.33227597145123e-05,
+      "loss": 1.0844,
+      "step": 25270
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3629555702209473,
+      "learning_rate": 4.3322495374041764e-05,
+      "loss": 1.3251,
+      "step": 25271
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8259289264678955,
+      "learning_rate": 4.3322231033571244e-05,
+      "loss": 1.4132,
+      "step": 25272
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.973227620124817,
+      "learning_rate": 4.332196669310072e-05,
+      "loss": 1.2209,
+      "step": 25273
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.788266658782959,
+      "learning_rate": 4.332170235263019e-05,
+      "loss": 1.1445,
+      "step": 25274
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.098698854446411,
+      "learning_rate": 4.3321438012159665e-05,
+      "loss": 1.3593,
+      "step": 25275
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.683603286743164,
+      "learning_rate": 4.332117367168914e-05,
+      "loss": 2.0884,
+      "step": 25276
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.106809616088867,
+      "learning_rate": 4.332090933121861e-05,
+      "loss": 1.2817,
+      "step": 25277
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.231513738632202,
+      "learning_rate": 4.3320644990748086e-05,
+      "loss": 1.4002,
+      "step": 25278
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.0294442176818848,
+      "learning_rate": 4.332038065027756e-05,
+      "loss": 1.2742,
+      "step": 25279
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.092811346054077,
+      "learning_rate": 4.332011630980703e-05,
+      "loss": 1.2364,
+      "step": 25280
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7835880517959595,
+      "learning_rate": 4.3319851969336507e-05,
+      "loss": 1.3371,
+      "step": 25281
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3162729740142822,
+      "learning_rate": 4.331958762886598e-05,
+      "loss": 1.4019,
+      "step": 25282
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6679099798202515,
+      "learning_rate": 4.331932328839546e-05,
+      "loss": 1.1876,
+      "step": 25283
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.46659517288208,
+      "learning_rate": 4.331905894792493e-05,
+      "loss": 1.4423,
+      "step": 25284
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1454145908355713,
+      "learning_rate": 4.331879460745441e-05,
+      "loss": 1.3579,
+      "step": 25285
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.000852584838867,
+      "learning_rate": 4.3318530266983874e-05,
+      "loss": 1.3655,
+      "step": 25286
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.101274251937866,
+      "learning_rate": 4.3318265926513355e-05,
+      "loss": 1.4068,
+      "step": 25287
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.896681547164917,
+      "learning_rate": 4.331800158604283e-05,
+      "loss": 1.2088,
+      "step": 25288
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8744778633117676,
+      "learning_rate": 4.33177372455723e-05,
+      "loss": 1.1647,
+      "step": 25289
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9766281843185425,
+      "learning_rate": 4.3317472905101775e-05,
+      "loss": 1.3304,
+      "step": 25290
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.760316014289856,
+      "learning_rate": 4.331720856463124e-05,
+      "loss": 1.3071,
+      "step": 25291
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2406489849090576,
+      "learning_rate": 4.331694422416072e-05,
+      "loss": 1.4078,
+      "step": 25292
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.994905948638916,
+      "learning_rate": 4.3316679883690196e-05,
+      "loss": 1.2417,
+      "step": 25293
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8829418420791626,
+      "learning_rate": 4.331641554321967e-05,
+      "loss": 1.255,
+      "step": 25294
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.898021936416626,
+      "learning_rate": 4.331615120274914e-05,
+      "loss": 1.3698,
+      "step": 25295
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.784287691116333,
+      "learning_rate": 4.331588686227862e-05,
+      "loss": 1.166,
+      "step": 25296
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9739716053009033,
+      "learning_rate": 4.331562252180809e-05,
+      "loss": 1.3008,
+      "step": 25297
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.921834111213684,
+      "learning_rate": 4.331535818133757e-05,
+      "loss": 1.2267,
+      "step": 25298
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.5514581203460693,
+      "learning_rate": 4.331509384086704e-05,
+      "loss": 1.3297,
+      "step": 25299
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.794506549835205,
+      "learning_rate": 4.331482950039651e-05,
+      "loss": 1.2605,
+      "step": 25300
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.4208483695983887,
+      "learning_rate": 4.3314565159925985e-05,
+      "loss": 1.4917,
+      "step": 25301
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.02972149848938,
+      "learning_rate": 4.331430081945546e-05,
+      "loss": 1.251,
+      "step": 25302
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.047236919403076,
+      "learning_rate": 4.331403647898494e-05,
+      "loss": 1.2925,
+      "step": 25303
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8397339582443237,
+      "learning_rate": 4.3313772138514405e-05,
+      "loss": 1.2376,
+      "step": 25304
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.4876000881195068,
+      "learning_rate": 4.3313507798043886e-05,
+      "loss": 1.1272,
+      "step": 25305
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.065891742706299,
+      "learning_rate": 4.331324345757335e-05,
+      "loss": 1.4138,
+      "step": 25306
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7857648134231567,
+      "learning_rate": 4.331297911710283e-05,
+      "loss": 1.4148,
+      "step": 25307
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6941572427749634,
+      "learning_rate": 4.3312714776632306e-05,
+      "loss": 1.2295,
+      "step": 25308
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.735989809036255,
+      "learning_rate": 4.331245043616178e-05,
+      "loss": 1.3931,
+      "step": 25309
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.5004303455352783,
+      "learning_rate": 4.3312186095691254e-05,
+      "loss": 1.3544,
+      "step": 25310
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.2014036178588867,
+      "learning_rate": 4.331192175522073e-05,
+      "loss": 1.546,
+      "step": 25311
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.06069278717041,
+      "learning_rate": 4.33116574147502e-05,
+      "loss": 1.2841,
+      "step": 25312
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6924690008163452,
+      "learning_rate": 4.3311393074279674e-05,
+      "loss": 1.3052,
+      "step": 25313
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.247947931289673,
+      "learning_rate": 4.331112873380915e-05,
+      "loss": 1.4298,
+      "step": 25314
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.654549241065979,
+      "learning_rate": 4.331086439333862e-05,
+      "loss": 1.1496,
+      "step": 25315
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6265478134155273,
+      "learning_rate": 4.3310600052868095e-05,
+      "loss": 1.1936,
+      "step": 25316
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9893039464950562,
+      "learning_rate": 4.331033571239757e-05,
+      "loss": 1.1866,
+      "step": 25317
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.4693444967269897,
+      "learning_rate": 4.331007137192705e-05,
+      "loss": 1.1579,
+      "step": 25318
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9233590364456177,
+      "learning_rate": 4.3309807031456516e-05,
+      "loss": 1.4367,
+      "step": 25319
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1182751655578613,
+      "learning_rate": 4.3309542690985996e-05,
+      "loss": 1.429,
+      "step": 25320
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1065640449523926,
+      "learning_rate": 4.330927835051546e-05,
+      "loss": 1.1853,
+      "step": 25321
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.263840913772583,
+      "learning_rate": 4.330901401004494e-05,
+      "loss": 1.4335,
+      "step": 25322
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8177270889282227,
+      "learning_rate": 4.330874966957442e-05,
+      "loss": 1.3505,
+      "step": 25323
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9904131889343262,
+      "learning_rate": 4.330848532910389e-05,
+      "loss": 1.286,
+      "step": 25324
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6969887018203735,
+      "learning_rate": 4.3308220988633364e-05,
+      "loss": 1.4091,
+      "step": 25325
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3129029273986816,
+      "learning_rate": 4.330795664816283e-05,
+      "loss": 1.495,
+      "step": 25326
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6409695148468018,
+      "learning_rate": 4.330769230769231e-05,
+      "loss": 1.2238,
+      "step": 25327
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7130720615386963,
+      "learning_rate": 4.3307427967221785e-05,
+      "loss": 1.226,
+      "step": 25328
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8120213747024536,
+      "learning_rate": 4.330716362675126e-05,
+      "loss": 1.2435,
+      "step": 25329
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.5708200931549072,
+      "learning_rate": 4.330689928628073e-05,
+      "loss": 1.4341,
+      "step": 25330
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.279085397720337,
+      "learning_rate": 4.3306634945810205e-05,
+      "loss": 1.4569,
+      "step": 25331
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.831180453300476,
+      "learning_rate": 4.330637060533968e-05,
+      "loss": 1.2234,
+      "step": 25332
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1328108310699463,
+      "learning_rate": 4.330610626486916e-05,
+      "loss": 1.4459,
+      "step": 25333
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.5474636554718018,
+      "learning_rate": 4.3305841924398626e-05,
+      "loss": 1.4454,
+      "step": 25334
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.356478691101074,
+      "learning_rate": 4.33055775839281e-05,
+      "loss": 1.3743,
+      "step": 25335
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2341291904449463,
+      "learning_rate": 4.330531324345757e-05,
+      "loss": 1.3398,
+      "step": 25336
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.8828723430633545,
+      "learning_rate": 4.330504890298705e-05,
+      "loss": 1.419,
+      "step": 25337
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7070547342300415,
+      "learning_rate": 4.330478456251653e-05,
+      "loss": 1.1523,
+      "step": 25338
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9105747938156128,
+      "learning_rate": 4.3304520222045994e-05,
+      "loss": 1.3236,
+      "step": 25339
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9800435304641724,
+      "learning_rate": 4.3304255881575474e-05,
+      "loss": 1.3311,
+      "step": 25340
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.4659860134124756,
+      "learning_rate": 4.330399154110494e-05,
+      "loss": 1.4181,
+      "step": 25341
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.0959665775299072,
+      "learning_rate": 4.330372720063442e-05,
+      "loss": 1.2882,
+      "step": 25342
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.5298893451690674,
+      "learning_rate": 4.3303462860163895e-05,
+      "loss": 1.4212,
+      "step": 25343
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1122329235076904,
+      "learning_rate": 4.330319851969337e-05,
+      "loss": 1.3155,
+      "step": 25344
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8720637559890747,
+      "learning_rate": 4.330293417922284e-05,
+      "loss": 1.3009,
+      "step": 25345
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.248986005783081,
+      "learning_rate": 4.3302669838752316e-05,
+      "loss": 1.3756,
+      "step": 25346
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2329416275024414,
+      "learning_rate": 4.330240549828179e-05,
+      "loss": 1.2368,
+      "step": 25347
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.7998619079589844,
+      "learning_rate": 4.330214115781126e-05,
+      "loss": 1.3782,
+      "step": 25348
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2397496700286865,
+      "learning_rate": 4.3301876817340736e-05,
+      "loss": 1.2876,
+      "step": 25349
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7028714418411255,
+      "learning_rate": 4.330161247687021e-05,
+      "loss": 1.1334,
+      "step": 25350
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9705455303192139,
+      "learning_rate": 4.3301348136399684e-05,
+      "loss": 1.2583,
+      "step": 25351
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.835383176803589,
+      "learning_rate": 4.330108379592916e-05,
+      "loss": 1.554,
+      "step": 25352
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.6865012645721436,
+      "learning_rate": 4.330081945545864e-05,
+      "loss": 1.5331,
+      "step": 25353
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.537505865097046,
+      "learning_rate": 4.3300555114988104e-05,
+      "loss": 1.3654,
+      "step": 25354
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7900452613830566,
+      "learning_rate": 4.3300290774517585e-05,
+      "loss": 1.3107,
+      "step": 25355
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.781782865524292,
+      "learning_rate": 4.330002643404705e-05,
+      "loss": 1.4403,
+      "step": 25356
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6782798767089844,
+      "learning_rate": 4.329976209357653e-05,
+      "loss": 1.0886,
+      "step": 25357
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.160717248916626,
+      "learning_rate": 4.3299497753106005e-05,
+      "loss": 1.4135,
+      "step": 25358
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2115254402160645,
+      "learning_rate": 4.329923341263548e-05,
+      "loss": 1.2514,
+      "step": 25359
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.60569429397583,
+      "learning_rate": 4.329896907216495e-05,
+      "loss": 1.2771,
+      "step": 25360
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9862303733825684,
+      "learning_rate": 4.3298704731694426e-05,
+      "loss": 1.3105,
+      "step": 25361
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.052229404449463,
+      "learning_rate": 4.32984403912239e-05,
+      "loss": 1.3189,
+      "step": 25362
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.4093000888824463,
+      "learning_rate": 4.329817605075337e-05,
+      "loss": 1.4045,
+      "step": 25363
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.229689359664917,
+      "learning_rate": 4.329791171028285e-05,
+      "loss": 1.3019,
+      "step": 25364
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1533305644989014,
+      "learning_rate": 4.329764736981232e-05,
+      "loss": 1.2705,
+      "step": 25365
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.5140109062194824,
+      "learning_rate": 4.3297383029341794e-05,
+      "loss": 1.2061,
+      "step": 25366
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.6175403594970703,
+      "learning_rate": 4.329711868887127e-05,
+      "loss": 1.5795,
+      "step": 25367
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8653595447540283,
+      "learning_rate": 4.329685434840075e-05,
+      "loss": 1.2487,
+      "step": 25368
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.0640273094177246,
+      "learning_rate": 4.3296590007930215e-05,
+      "loss": 1.8271,
+      "step": 25369
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.8185014724731445,
+      "learning_rate": 4.3296325667459695e-05,
+      "loss": 1.3731,
+      "step": 25370
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.4440503120422363,
+      "learning_rate": 4.329606132698916e-05,
+      "loss": 1.2701,
+      "step": 25371
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1476407051086426,
+      "learning_rate": 4.3295796986518635e-05,
+      "loss": 1.3291,
+      "step": 25372
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7580209970474243,
+      "learning_rate": 4.3295532646048116e-05,
+      "loss": 1.2969,
+      "step": 25373
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.656308650970459,
+      "learning_rate": 4.329526830557758e-05,
+      "loss": 1.4309,
+      "step": 25374
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.893578290939331,
+      "learning_rate": 4.329500396510706e-05,
+      "loss": 1.3481,
+      "step": 25375
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3620669841766357,
+      "learning_rate": 4.329473962463653e-05,
+      "loss": 1.5873,
+      "step": 25376
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.5939557552337646,
+      "learning_rate": 4.329447528416601e-05,
+      "loss": 1.2007,
+      "step": 25377
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9025799036026,
+      "learning_rate": 4.3294210943695484e-05,
+      "loss": 1.3407,
+      "step": 25378
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2340431213378906,
+      "learning_rate": 4.329394660322496e-05,
+      "loss": 1.3463,
+      "step": 25379
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.833869457244873,
+      "learning_rate": 4.329368226275443e-05,
+      "loss": 1.3113,
+      "step": 25380
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2794127464294434,
+      "learning_rate": 4.3293417922283904e-05,
+      "loss": 1.2438,
+      "step": 25381
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.8683340549468994,
+      "learning_rate": 4.329315358181338e-05,
+      "loss": 1.3525,
+      "step": 25382
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7984256744384766,
+      "learning_rate": 4.329288924134285e-05,
+      "loss": 1.2545,
+      "step": 25383
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8165990114212036,
+      "learning_rate": 4.3292624900872325e-05,
+      "loss": 1.2402,
+      "step": 25384
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8052600622177124,
+      "learning_rate": 4.32923605604018e-05,
+      "loss": 1.4014,
+      "step": 25385
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8007274866104126,
+      "learning_rate": 4.329209621993127e-05,
+      "loss": 1.2706,
+      "step": 25386
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.6402382850646973,
+      "learning_rate": 4.3291831879460746e-05,
+      "loss": 1.3533,
+      "step": 25387
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6873254776000977,
+      "learning_rate": 4.3291567538990226e-05,
+      "loss": 1.218,
+      "step": 25388
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3169069290161133,
+      "learning_rate": 4.329130319851969e-05,
+      "loss": 1.1464,
+      "step": 25389
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.0299901962280273,
+      "learning_rate": 4.329103885804917e-05,
+      "loss": 1.0582,
+      "step": 25390
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.1199138164520264,
+      "learning_rate": 4.329077451757864e-05,
+      "loss": 1.5384,
+      "step": 25391
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8106578588485718,
+      "learning_rate": 4.329051017710812e-05,
+      "loss": 1.0674,
+      "step": 25392
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.559821605682373,
+      "learning_rate": 4.3290245836637594e-05,
+      "loss": 1.365,
+      "step": 25393
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.5334386825561523,
+      "learning_rate": 4.328998149616707e-05,
+      "loss": 1.597,
+      "step": 25394
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.085620641708374,
+      "learning_rate": 4.328971715569654e-05,
+      "loss": 1.2759,
+      "step": 25395
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3051140308380127,
+      "learning_rate": 4.3289452815226015e-05,
+      "loss": 1.3548,
+      "step": 25396
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.0259358882904053,
+      "learning_rate": 4.328918847475549e-05,
+      "loss": 1.4057,
+      "step": 25397
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9425126314163208,
+      "learning_rate": 4.328892413428496e-05,
+      "loss": 1.2607,
+      "step": 25398
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6825307607650757,
+      "learning_rate": 4.3288659793814435e-05,
+      "loss": 1.2046,
+      "step": 25399
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3617732524871826,
+      "learning_rate": 4.328839545334391e-05,
+      "loss": 1.3726,
+      "step": 25400
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9931975603103638,
+      "learning_rate": 4.328813111287338e-05,
+      "loss": 1.2739,
+      "step": 25401
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.179332971572876,
+      "learning_rate": 4.3287866772402856e-05,
+      "loss": 1.3644,
+      "step": 25402
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.750521421432495,
+      "learning_rate": 4.3287602431932336e-05,
+      "loss": 1.4453,
+      "step": 25403
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.559870958328247,
+      "learning_rate": 4.32873380914618e-05,
+      "loss": 1.2501,
+      "step": 25404
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9522138833999634,
+      "learning_rate": 4.3287073750991284e-05,
+      "loss": 1.1757,
+      "step": 25405
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.84699547290802,
+      "learning_rate": 4.328680941052075e-05,
+      "loss": 1.2384,
+      "step": 25406
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.631081461906433,
+      "learning_rate": 4.328654507005023e-05,
+      "loss": 1.1625,
+      "step": 25407
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.4269580841064453,
+      "learning_rate": 4.3286280729579704e-05,
+      "loss": 1.2692,
+      "step": 25408
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6115827560424805,
+      "learning_rate": 4.328601638910917e-05,
+      "loss": 1.1071,
+      "step": 25409
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.655486583709717,
+      "learning_rate": 4.328575204863865e-05,
+      "loss": 1.312,
+      "step": 25410
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1363351345062256,
+      "learning_rate": 4.328548770816812e-05,
+      "loss": 1.289,
+      "step": 25411
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9615546464920044,
+      "learning_rate": 4.32852233676976e-05,
+      "loss": 1.4539,
+      "step": 25412
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7427819967269897,
+      "learning_rate": 4.328495902722707e-05,
+      "loss": 1.2059,
+      "step": 25413
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.51286244392395,
+      "learning_rate": 4.3284694686756546e-05,
+      "loss": 1.3345,
+      "step": 25414
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3600120544433594,
+      "learning_rate": 4.328443034628602e-05,
+      "loss": 1.4905,
+      "step": 25415
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.870748996734619,
+      "learning_rate": 4.328416600581549e-05,
+      "loss": 1.4918,
+      "step": 25416
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9512193202972412,
+      "learning_rate": 4.3283901665344966e-05,
+      "loss": 1.2693,
+      "step": 25417
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2387924194335938,
+      "learning_rate": 4.328363732487444e-05,
+      "loss": 1.1955,
+      "step": 25418
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7102888822555542,
+      "learning_rate": 4.3283372984403913e-05,
+      "loss": 1.168,
+      "step": 25419
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.08725643157959,
+      "learning_rate": 4.328310864393339e-05,
+      "loss": 1.3381,
+      "step": 25420
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.117886781692505,
+      "learning_rate": 4.328284430346286e-05,
+      "loss": 1.1787,
+      "step": 25421
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.847174882888794,
+      "learning_rate": 4.3282579962992334e-05,
+      "loss": 1.3797,
+      "step": 25422
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.037266969680786,
+      "learning_rate": 4.3282315622521815e-05,
+      "loss": 1.2471,
+      "step": 25423
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8280283212661743,
+      "learning_rate": 4.328205128205128e-05,
+      "loss": 1.2498,
+      "step": 25424
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.5804153680801392,
+      "learning_rate": 4.328178694158076e-05,
+      "loss": 1.1412,
+      "step": 25425
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9650681018829346,
+      "learning_rate": 4.328152260111023e-05,
+      "loss": 1.2766,
+      "step": 25426
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8871688842773438,
+      "learning_rate": 4.328125826063971e-05,
+      "loss": 1.3381,
+      "step": 25427
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8263702392578125,
+      "learning_rate": 4.3280993920169176e-05,
+      "loss": 1.1892,
+      "step": 25428
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.072784423828125,
+      "learning_rate": 4.3280729579698656e-05,
+      "loss": 1.2454,
+      "step": 25429
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.0125484466552734,
+      "learning_rate": 4.328046523922813e-05,
+      "loss": 1.2747,
+      "step": 25430
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.488279342651367,
+      "learning_rate": 4.32802008987576e-05,
+      "loss": 1.3009,
+      "step": 25431
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.470008134841919,
+      "learning_rate": 4.327993655828708e-05,
+      "loss": 1.437,
+      "step": 25432
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1908669471740723,
+      "learning_rate": 4.327967221781655e-05,
+      "loss": 1.2779,
+      "step": 25433
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1081931591033936,
+      "learning_rate": 4.3279407877346024e-05,
+      "loss": 1.291,
+      "step": 25434
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8357104063034058,
+      "learning_rate": 4.32791435368755e-05,
+      "loss": 1.1939,
+      "step": 25435
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1966605186462402,
+      "learning_rate": 4.327887919640497e-05,
+      "loss": 1.305,
+      "step": 25436
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.809901237487793,
+      "learning_rate": 4.3278614855934445e-05,
+      "loss": 1.2149,
+      "step": 25437
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.444478750228882,
+      "learning_rate": 4.327835051546392e-05,
+      "loss": 1.3846,
+      "step": 25438
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.468320846557617,
+      "learning_rate": 4.327808617499339e-05,
+      "loss": 1.1706,
+      "step": 25439
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9578872919082642,
+      "learning_rate": 4.327782183452287e-05,
+      "loss": 1.2504,
+      "step": 25440
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7456085681915283,
+      "learning_rate": 4.327755749405234e-05,
+      "loss": 2.1126,
+      "step": 25441
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.246722936630249,
+      "learning_rate": 4.327729315358182e-05,
+      "loss": 1.4334,
+      "step": 25442
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0115933418273926,
+      "learning_rate": 4.3277028813111286e-05,
+      "loss": 1.2284,
+      "step": 25443
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9725078344345093,
+      "learning_rate": 4.327676447264076e-05,
+      "loss": 2.3561,
+      "step": 25444
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9002177715301514,
+      "learning_rate": 4.327650013217024e-05,
+      "loss": 1.2825,
+      "step": 25445
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1578423976898193,
+      "learning_rate": 4.327623579169971e-05,
+      "loss": 1.2919,
+      "step": 25446
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1349802017211914,
+      "learning_rate": 4.327597145122919e-05,
+      "loss": 1.243,
+      "step": 25447
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7137322425842285,
+      "learning_rate": 4.3275707110758654e-05,
+      "loss": 1.35,
+      "step": 25448
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.5362118482589722,
+      "learning_rate": 4.3275442770288134e-05,
+      "loss": 1.1616,
+      "step": 25449
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.559964656829834,
+      "learning_rate": 4.327517842981761e-05,
+      "loss": 1.282,
+      "step": 25450
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7291892766952515,
+      "learning_rate": 4.327491408934708e-05,
+      "loss": 1.2467,
+      "step": 25451
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1478846073150635,
+      "learning_rate": 4.3274649748876555e-05,
+      "loss": 1.3928,
+      "step": 25452
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.786149740219116,
+      "learning_rate": 4.327438540840603e-05,
+      "loss": 1.4712,
+      "step": 25453
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.00225567817688,
+      "learning_rate": 4.32741210679355e-05,
+      "loss": 1.2011,
+      "step": 25454
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.545057535171509,
+      "learning_rate": 4.3273856727464976e-05,
+      "loss": 1.3965,
+      "step": 25455
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8678317070007324,
+      "learning_rate": 4.327359238699445e-05,
+      "loss": 1.4673,
+      "step": 25456
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1142489910125732,
+      "learning_rate": 4.327332804652392e-05,
+      "loss": 1.4099,
+      "step": 25457
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0446205139160156,
+      "learning_rate": 4.3273063706053396e-05,
+      "loss": 1.3296,
+      "step": 25458
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0737013816833496,
+      "learning_rate": 4.327279936558287e-05,
+      "loss": 1.4044,
+      "step": 25459
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1695046424865723,
+      "learning_rate": 4.327253502511235e-05,
+      "loss": 1.3766,
+      "step": 25460
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1983835697174072,
+      "learning_rate": 4.327227068464182e-05,
+      "loss": 1.3411,
+      "step": 25461
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.062812566757202,
+      "learning_rate": 4.32720063441713e-05,
+      "loss": 1.5567,
+      "step": 25462
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9173210859298706,
+      "learning_rate": 4.3271742003700764e-05,
+      "loss": 1.3496,
+      "step": 25463
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9240562915802002,
+      "learning_rate": 4.3271477663230245e-05,
+      "loss": 1.2166,
+      "step": 25464
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.4055187702178955,
+      "learning_rate": 4.327121332275972e-05,
+      "loss": 1.5125,
+      "step": 25465
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.852691888809204,
+      "learning_rate": 4.327094898228919e-05,
+      "loss": 1.3199,
+      "step": 25466
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.705482006072998,
+      "learning_rate": 4.3270684641818665e-05,
+      "loss": 1.2021,
+      "step": 25467
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.512753486633301,
+      "learning_rate": 4.327042030134814e-05,
+      "loss": 1.3216,
+      "step": 25468
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1771347522735596,
+      "learning_rate": 4.327015596087761e-05,
+      "loss": 1.243,
+      "step": 25469
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9550353288650513,
+      "learning_rate": 4.3269891620407086e-05,
+      "loss": 1.2873,
+      "step": 25470
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9592872858047485,
+      "learning_rate": 4.326962727993656e-05,
+      "loss": 1.1666,
+      "step": 25471
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0924582481384277,
+      "learning_rate": 4.326936293946603e-05,
+      "loss": 1.3837,
+      "step": 25472
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.146172285079956,
+      "learning_rate": 4.326909859899551e-05,
+      "loss": 1.1545,
+      "step": 25473
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.243594169616699,
+      "learning_rate": 4.326883425852498e-05,
+      "loss": 1.1834,
+      "step": 25474
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7757433652877808,
+      "learning_rate": 4.326856991805446e-05,
+      "loss": 1.2186,
+      "step": 25475
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.87405264377594,
+      "learning_rate": 4.326830557758393e-05,
+      "loss": 1.3905,
+      "step": 25476
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.118809938430786,
+      "learning_rate": 4.326804123711341e-05,
+      "loss": 1.0491,
+      "step": 25477
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1418797969818115,
+      "learning_rate": 4.3267776896642875e-05,
+      "loss": 1.4944,
+      "step": 25478
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0107905864715576,
+      "learning_rate": 4.3267512556172355e-05,
+      "loss": 1.2976,
+      "step": 25479
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.032661199569702,
+      "learning_rate": 4.326724821570183e-05,
+      "loss": 1.2065,
+      "step": 25480
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9932032823562622,
+      "learning_rate": 4.3266983875231295e-05,
+      "loss": 1.2632,
+      "step": 25481
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.282261371612549,
+      "learning_rate": 4.3266719534760776e-05,
+      "loss": 1.4202,
+      "step": 25482
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.275667667388916,
+      "learning_rate": 4.326645519429024e-05,
+      "loss": 1.1985,
+      "step": 25483
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.116326093673706,
+      "learning_rate": 4.326619085381972e-05,
+      "loss": 1.3917,
+      "step": 25484
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7496979236602783,
+      "learning_rate": 4.3265926513349196e-05,
+      "loss": 1.223,
+      "step": 25485
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1991708278656006,
+      "learning_rate": 4.326566217287867e-05,
+      "loss": 1.2825,
+      "step": 25486
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6515103578567505,
+      "learning_rate": 4.3265397832408143e-05,
+      "loss": 1.187,
+      "step": 25487
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.633925199508667,
+      "learning_rate": 4.326513349193762e-05,
+      "loss": 1.2661,
+      "step": 25488
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.335373640060425,
+      "learning_rate": 4.326486915146709e-05,
+      "loss": 1.4464,
+      "step": 25489
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9086499214172363,
+      "learning_rate": 4.3264604810996564e-05,
+      "loss": 1.3291,
+      "step": 25490
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.2193429470062256,
+      "learning_rate": 4.326434047052604e-05,
+      "loss": 1.3141,
+      "step": 25491
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.11496639251709,
+      "learning_rate": 4.326407613005551e-05,
+      "loss": 1.4187,
+      "step": 25492
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7845582962036133,
+      "learning_rate": 4.3263811789584985e-05,
+      "loss": 1.2844,
+      "step": 25493
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.5285236835479736,
+      "learning_rate": 4.326354744911446e-05,
+      "loss": 1.1539,
+      "step": 25494
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.933099627494812,
+      "learning_rate": 4.326328310864394e-05,
+      "loss": 1.2138,
+      "step": 25495
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6861127614974976,
+      "learning_rate": 4.3263018768173406e-05,
+      "loss": 1.2213,
+      "step": 25496
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7937021255493164,
+      "learning_rate": 4.3262754427702886e-05,
+      "loss": 1.2419,
+      "step": 25497
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.651831865310669,
+      "learning_rate": 4.326249008723235e-05,
+      "loss": 1.1354,
+      "step": 25498
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.935973882675171,
+      "learning_rate": 4.326222574676183e-05,
+      "loss": 1.3698,
+      "step": 25499
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.884116291999817,
+      "learning_rate": 4.326196140629131e-05,
+      "loss": 1.3707,
+      "step": 25500
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.2287395000457764,
+      "learning_rate": 4.326169706582078e-05,
+      "loss": 1.1628,
+      "step": 25501
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1259658336639404,
+      "learning_rate": 4.3261432725350254e-05,
+      "loss": 1.203,
+      "step": 25502
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.709001898765564,
+      "learning_rate": 4.326116838487973e-05,
+      "loss": 1.343,
+      "step": 25503
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.2365357875823975,
+      "learning_rate": 4.32609040444092e-05,
+      "loss": 1.4635,
+      "step": 25504
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.12485671043396,
+      "learning_rate": 4.3260639703938674e-05,
+      "loss": 1.2879,
+      "step": 25505
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7285033464431763,
+      "learning_rate": 4.326037536346815e-05,
+      "loss": 1.2191,
+      "step": 25506
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0997955799102783,
+      "learning_rate": 4.326011102299762e-05,
+      "loss": 1.3116,
+      "step": 25507
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.137338161468506,
+      "learning_rate": 4.3259846682527095e-05,
+      "loss": 1.5346,
+      "step": 25508
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7486729621887207,
+      "learning_rate": 4.325958234205657e-05,
+      "loss": 1.1329,
+      "step": 25509
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1198670864105225,
+      "learning_rate": 4.325931800158605e-05,
+      "loss": 1.8935,
+      "step": 25510
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.199388265609741,
+      "learning_rate": 4.3259053661115516e-05,
+      "loss": 1.4361,
+      "step": 25511
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.2193446159362793,
+      "learning_rate": 4.3258789320644996e-05,
+      "loss": 1.4047,
+      "step": 25512
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8676058053970337,
+      "learning_rate": 4.325852498017446e-05,
+      "loss": 1.2803,
+      "step": 25513
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.2324609756469727,
+      "learning_rate": 4.3258260639703943e-05,
+      "loss": 1.2536,
+      "step": 25514
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.5899055004119873,
+      "learning_rate": 4.325799629923342e-05,
+      "loss": 1.6046,
+      "step": 25515
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9894630908966064,
+      "learning_rate": 4.3257731958762884e-05,
+      "loss": 1.4025,
+      "step": 25516
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.101675510406494,
+      "learning_rate": 4.3257467618292364e-05,
+      "loss": 1.2537,
+      "step": 25517
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9261137247085571,
+      "learning_rate": 4.325720327782183e-05,
+      "loss": 1.3345,
+      "step": 25518
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0964279174804688,
+      "learning_rate": 4.325693893735131e-05,
+      "loss": 1.3259,
+      "step": 25519
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0703461170196533,
+      "learning_rate": 4.3256674596880785e-05,
+      "loss": 1.2327,
+      "step": 25520
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8691534996032715,
+      "learning_rate": 4.325641025641026e-05,
+      "loss": 2.0267,
+      "step": 25521
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.6894240379333496,
+      "learning_rate": 4.325614591593973e-05,
+      "loss": 1.3841,
+      "step": 25522
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.5661768913269043,
+      "learning_rate": 4.3255881575469206e-05,
+      "loss": 1.3965,
+      "step": 25523
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9995654821395874,
+      "learning_rate": 4.325561723499868e-05,
+      "loss": 1.3131,
+      "step": 25524
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0301105976104736,
+      "learning_rate": 4.325535289452816e-05,
+      "loss": 1.3007,
+      "step": 25525
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9789390563964844,
+      "learning_rate": 4.3255088554057626e-05,
+      "loss": 1.322,
+      "step": 25526
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.5649335384368896,
+      "learning_rate": 4.32548242135871e-05,
+      "loss": 1.5475,
+      "step": 25527
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8634655475616455,
+      "learning_rate": 4.325455987311657e-05,
+      "loss": 1.2517,
+      "step": 25528
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8500874042510986,
+      "learning_rate": 4.325429553264605e-05,
+      "loss": 1.267,
+      "step": 25529
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7462869882583618,
+      "learning_rate": 4.325403119217553e-05,
+      "loss": 1.3443,
+      "step": 25530
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6025136709213257,
+      "learning_rate": 4.3253766851704994e-05,
+      "loss": 1.16,
+      "step": 25531
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1396028995513916,
+      "learning_rate": 4.3253502511234474e-05,
+      "loss": 1.3148,
+      "step": 25532
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1248624324798584,
+      "learning_rate": 4.325323817076394e-05,
+      "loss": 1.3964,
+      "step": 25533
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6727784872055054,
+      "learning_rate": 4.325297383029342e-05,
+      "loss": 1.6057,
+      "step": 25534
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8814010620117188,
+      "learning_rate": 4.3252709489822895e-05,
+      "loss": 1.2268,
+      "step": 25535
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.389347791671753,
+      "learning_rate": 4.325244514935237e-05,
+      "loss": 1.3702,
+      "step": 25536
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1739349365234375,
+      "learning_rate": 4.325218080888184e-05,
+      "loss": 1.3255,
+      "step": 25537
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9173743724822998,
+      "learning_rate": 4.3251916468411316e-05,
+      "loss": 1.1402,
+      "step": 25538
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8935585021972656,
+      "learning_rate": 4.325165212794079e-05,
+      "loss": 1.2016,
+      "step": 25539
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.349543333053589,
+      "learning_rate": 4.325138778747026e-05,
+      "loss": 2.2623,
+      "step": 25540
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.255585193634033,
+      "learning_rate": 4.3251123446999737e-05,
+      "loss": 1.3964,
+      "step": 25541
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.6653876304626465,
+      "learning_rate": 4.325085910652921e-05,
+      "loss": 1.3191,
+      "step": 25542
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.2453179359436035,
+      "learning_rate": 4.3250594766058684e-05,
+      "loss": 1.3968,
+      "step": 25543
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.187535524368286,
+      "learning_rate": 4.325033042558816e-05,
+      "loss": 1.3518,
+      "step": 25544
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.861108422279358,
+      "learning_rate": 4.325006608511764e-05,
+      "loss": 1.2806,
+      "step": 25545
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8755316734313965,
+      "learning_rate": 4.3249801744647104e-05,
+      "loss": 1.2668,
+      "step": 25546
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6924598217010498,
+      "learning_rate": 4.3249537404176585e-05,
+      "loss": 1.2421,
+      "step": 25547
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.090475559234619,
+      "learning_rate": 4.324927306370605e-05,
+      "loss": 1.4754,
+      "step": 25548
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.141939878463745,
+      "learning_rate": 4.324900872323553e-05,
+      "loss": 1.353,
+      "step": 25549
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3220560550689697,
+      "learning_rate": 4.3248744382765005e-05,
+      "loss": 1.3462,
+      "step": 25550
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.164862871170044,
+      "learning_rate": 4.324848004229448e-05,
+      "loss": 1.4687,
+      "step": 25551
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3972926139831543,
+      "learning_rate": 4.324821570182395e-05,
+      "loss": 1.2676,
+      "step": 25552
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.436979055404663,
+      "learning_rate": 4.324795136135342e-05,
+      "loss": 1.5413,
+      "step": 25553
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.4910197257995605,
+      "learning_rate": 4.32476870208829e-05,
+      "loss": 1.5196,
+      "step": 25554
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8109767436981201,
+      "learning_rate": 4.324742268041237e-05,
+      "loss": 1.149,
+      "step": 25555
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.107097864151001,
+      "learning_rate": 4.324715833994185e-05,
+      "loss": 1.2708,
+      "step": 25556
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.185546636581421,
+      "learning_rate": 4.324689399947132e-05,
+      "loss": 1.3134,
+      "step": 25557
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.67279314994812,
+      "learning_rate": 4.3246629659000794e-05,
+      "loss": 1.3458,
+      "step": 25558
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3564202785491943,
+      "learning_rate": 4.324636531853027e-05,
+      "loss": 1.3301,
+      "step": 25559
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9898059368133545,
+      "learning_rate": 4.324610097805975e-05,
+      "loss": 1.2168,
+      "step": 25560
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6874929666519165,
+      "learning_rate": 4.3245836637589215e-05,
+      "loss": 1.2462,
+      "step": 25561
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.115943193435669,
+      "learning_rate": 4.324557229711869e-05,
+      "loss": 1.1694,
+      "step": 25562
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.863549828529358,
+      "learning_rate": 4.324530795664816e-05,
+      "loss": 1.1915,
+      "step": 25563
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.6290152072906494,
+      "learning_rate": 4.3245043616177635e-05,
+      "loss": 1.4488,
+      "step": 25564
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.784970998764038,
+      "learning_rate": 4.3244779275707116e-05,
+      "loss": 1.2033,
+      "step": 25565
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.817156195640564,
+      "learning_rate": 4.324451493523658e-05,
+      "loss": 1.3013,
+      "step": 25566
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3419766426086426,
+      "learning_rate": 4.324425059476606e-05,
+      "loss": 1.3315,
+      "step": 25567
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6755818128585815,
+      "learning_rate": 4.324398625429553e-05,
+      "loss": 1.3088,
+      "step": 25568
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.296128749847412,
+      "learning_rate": 4.324372191382501e-05,
+      "loss": 1.216,
+      "step": 25569
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.029712200164795,
+      "learning_rate": 4.3243457573354484e-05,
+      "loss": 1.3529,
+      "step": 25570
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9611233472824097,
+      "learning_rate": 4.324319323288396e-05,
+      "loss": 1.3741,
+      "step": 25571
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.539504289627075,
+      "learning_rate": 4.324292889241343e-05,
+      "loss": 1.2758,
+      "step": 25572
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3236114978790283,
+      "learning_rate": 4.3242664551942904e-05,
+      "loss": 1.3482,
+      "step": 25573
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.8742072582244873,
+      "learning_rate": 4.324240021147238e-05,
+      "loss": 1.9558,
+      "step": 25574
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.928445816040039,
+      "learning_rate": 4.324213587100185e-05,
+      "loss": 1.2728,
+      "step": 25575
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.086801290512085,
+      "learning_rate": 4.3241871530531325e-05,
+      "loss": 1.2546,
+      "step": 25576
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9843857288360596,
+      "learning_rate": 4.32416071900608e-05,
+      "loss": 1.3466,
+      "step": 25577
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9901214838027954,
+      "learning_rate": 4.324134284959027e-05,
+      "loss": 1.6528,
+      "step": 25578
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3269245624542236,
+      "learning_rate": 4.3241078509119746e-05,
+      "loss": 1.4211,
+      "step": 25579
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3472068309783936,
+      "learning_rate": 4.3240814168649226e-05,
+      "loss": 1.2721,
+      "step": 25580
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6916548013687134,
+      "learning_rate": 4.324054982817869e-05,
+      "loss": 1.3517,
+      "step": 25581
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.207029342651367,
+      "learning_rate": 4.324028548770817e-05,
+      "loss": 1.3489,
+      "step": 25582
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6902036666870117,
+      "learning_rate": 4.324002114723764e-05,
+      "loss": 2.3171,
+      "step": 25583
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9482015371322632,
+      "learning_rate": 4.323975680676712e-05,
+      "loss": 1.2077,
+      "step": 25584
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.165703296661377,
+      "learning_rate": 4.3239492466296594e-05,
+      "loss": 1.2547,
+      "step": 25585
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.657753586769104,
+      "learning_rate": 4.323922812582607e-05,
+      "loss": 1.1669,
+      "step": 25586
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.713421106338501,
+      "learning_rate": 4.323896378535554e-05,
+      "loss": 1.2909,
+      "step": 25587
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8262866735458374,
+      "learning_rate": 4.323869944488501e-05,
+      "loss": 1.3315,
+      "step": 25588
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.893261194229126,
+      "learning_rate": 4.323843510441449e-05,
+      "loss": 1.241,
+      "step": 25589
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.145911931991577,
+      "learning_rate": 4.323817076394396e-05,
+      "loss": 1.3923,
+      "step": 25590
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3850209712982178,
+      "learning_rate": 4.3237906423473435e-05,
+      "loss": 1.309,
+      "step": 25591
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.016180992126465,
+      "learning_rate": 4.323764208300291e-05,
+      "loss": 1.5752,
+      "step": 25592
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.046727180480957,
+      "learning_rate": 4.323737774253238e-05,
+      "loss": 1.3728,
+      "step": 25593
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6313331127166748,
+      "learning_rate": 4.3237113402061856e-05,
+      "loss": 1.3674,
+      "step": 25594
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.8828375339508057,
+      "learning_rate": 4.3236849061591337e-05,
+      "loss": 1.343,
+      "step": 25595
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.494354724884033,
+      "learning_rate": 4.32365847211208e-05,
+      "loss": 1.4623,
+      "step": 25596
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.2176387310028076,
+      "learning_rate": 4.3236320380650284e-05,
+      "loss": 1.5558,
+      "step": 25597
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8342070579528809,
+      "learning_rate": 4.323605604017975e-05,
+      "loss": 1.1475,
+      "step": 25598
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8944544792175293,
+      "learning_rate": 4.3235791699709224e-05,
+      "loss": 1.316,
+      "step": 25599
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.108856678009033,
+      "learning_rate": 4.3235527359238704e-05,
+      "loss": 1.2799,
+      "step": 25600
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1555936336517334,
+      "learning_rate": 4.323526301876817e-05,
+      "loss": 1.3845,
+      "step": 25601
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.084068775177002,
+      "learning_rate": 4.323499867829765e-05,
+      "loss": 1.4256,
+      "step": 25602
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8669196367263794,
+      "learning_rate": 4.323473433782712e-05,
+      "loss": 1.3181,
+      "step": 25603
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.374485492706299,
+      "learning_rate": 4.32344699973566e-05,
+      "loss": 1.3542,
+      "step": 25604
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9851912260055542,
+      "learning_rate": 4.323420565688607e-05,
+      "loss": 1.281,
+      "step": 25605
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8056117296218872,
+      "learning_rate": 4.3233941316415546e-05,
+      "loss": 1.2613,
+      "step": 25606
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.2371208667755127,
+      "learning_rate": 4.323367697594502e-05,
+      "loss": 1.354,
+      "step": 25607
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0249319076538086,
+      "learning_rate": 4.323341263547449e-05,
+      "loss": 1.1933,
+      "step": 25608
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9405914545059204,
+      "learning_rate": 4.3233148295003967e-05,
+      "loss": 1.2395,
+      "step": 25609
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8157678842544556,
+      "learning_rate": 4.323288395453344e-05,
+      "loss": 2.2142,
+      "step": 25610
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1669955253601074,
+      "learning_rate": 4.3232619614062914e-05,
+      "loss": 2.4889,
+      "step": 25611
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.724949836730957,
+      "learning_rate": 4.323235527359239e-05,
+      "loss": 1.5137,
+      "step": 25612
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1407394409179688,
+      "learning_rate": 4.323209093312186e-05,
+      "loss": 1.1365,
+      "step": 25613
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9859137535095215,
+      "learning_rate": 4.3231826592651334e-05,
+      "loss": 1.2486,
+      "step": 25614
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8167515993118286,
+      "learning_rate": 4.3231562252180815e-05,
+      "loss": 1.1676,
+      "step": 25615
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.988153338432312,
+      "learning_rate": 4.323129791171028e-05,
+      "loss": 1.1959,
+      "step": 25616
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9779044389724731,
+      "learning_rate": 4.323103357123976e-05,
+      "loss": 1.3453,
+      "step": 25617
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.0372817516326904,
+      "learning_rate": 4.323076923076923e-05,
+      "loss": 1.2135,
+      "step": 25618
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.882432222366333,
+      "learning_rate": 4.323050489029871e-05,
+      "loss": 1.5797,
+      "step": 25619
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7505569458007812,
+      "learning_rate": 4.323024054982818e-05,
+      "loss": 1.1369,
+      "step": 25620
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.805180311203003,
+      "learning_rate": 4.3229976209357656e-05,
+      "loss": 1.1965,
+      "step": 25621
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8105456829071045,
+      "learning_rate": 4.322971186888713e-05,
+      "loss": 1.2454,
+      "step": 25622
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1404306888580322,
+      "learning_rate": 4.32294475284166e-05,
+      "loss": 1.2549,
+      "step": 25623
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.318209409713745,
+      "learning_rate": 4.322918318794608e-05,
+      "loss": 1.2995,
+      "step": 25624
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.067919969558716,
+      "learning_rate": 4.322891884747555e-05,
+      "loss": 1.3859,
+      "step": 25625
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0541160106658936,
+      "learning_rate": 4.3228654507005024e-05,
+      "loss": 1.1589,
+      "step": 25626
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.4274988174438477,
+      "learning_rate": 4.32283901665345e-05,
+      "loss": 1.3648,
+      "step": 25627
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.184314727783203,
+      "learning_rate": 4.322812582606397e-05,
+      "loss": 1.2606,
+      "step": 25628
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3747546672821045,
+      "learning_rate": 4.3227861485593445e-05,
+      "loss": 1.2722,
+      "step": 25629
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6548552513122559,
+      "learning_rate": 4.3227597145122925e-05,
+      "loss": 1.0906,
+      "step": 25630
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7949297428131104,
+      "learning_rate": 4.322733280465239e-05,
+      "loss": 1.1615,
+      "step": 25631
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.529143810272217,
+      "learning_rate": 4.322706846418187e-05,
+      "loss": 1.5541,
+      "step": 25632
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9036976099014282,
+      "learning_rate": 4.322680412371134e-05,
+      "loss": 2.2471,
+      "step": 25633
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.73920738697052,
+      "learning_rate": 4.322653978324081e-05,
+      "loss": 1.1986,
+      "step": 25634
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0501694679260254,
+      "learning_rate": 4.322627544277029e-05,
+      "loss": 1.3354,
+      "step": 25635
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0072147846221924,
+      "learning_rate": 4.322601110229976e-05,
+      "loss": 1.1325,
+      "step": 25636
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.6324347257614136,
+      "learning_rate": 4.322574676182924e-05,
+      "loss": 2.4178,
+      "step": 25637
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.205195426940918,
+      "learning_rate": 4.322548242135871e-05,
+      "loss": 1.4001,
+      "step": 25638
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.709581136703491,
+      "learning_rate": 4.322521808088819e-05,
+      "loss": 1.263,
+      "step": 25639
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2039480209350586,
+      "learning_rate": 4.322495374041766e-05,
+      "loss": 1.234,
+      "step": 25640
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2479593753814697,
+      "learning_rate": 4.3224689399947134e-05,
+      "loss": 1.4025,
+      "step": 25641
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.853677988052368,
+      "learning_rate": 4.322442505947661e-05,
+      "loss": 1.5037,
+      "step": 25642
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8573737144470215,
+      "learning_rate": 4.322416071900608e-05,
+      "loss": 1.2201,
+      "step": 25643
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.6174144744873047,
+      "learning_rate": 4.3223896378535555e-05,
+      "loss": 1.2957,
+      "step": 25644
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.535623550415039,
+      "learning_rate": 4.322363203806503e-05,
+      "loss": 1.435,
+      "step": 25645
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.072319507598877,
+      "learning_rate": 4.32233676975945e-05,
+      "loss": 1.3234,
+      "step": 25646
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8885377645492554,
+      "learning_rate": 4.3223103357123976e-05,
+      "loss": 1.2554,
+      "step": 25647
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.878706455230713,
+      "learning_rate": 4.322283901665345e-05,
+      "loss": 1.3611,
+      "step": 25648
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.4570579528808594,
+      "learning_rate": 4.322257467618292e-05,
+      "loss": 1.5165,
+      "step": 25649
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.168294906616211,
+      "learning_rate": 4.32223103357124e-05,
+      "loss": 2.4635,
+      "step": 25650
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.835334539413452,
+      "learning_rate": 4.322204599524187e-05,
+      "loss": 1.4686,
+      "step": 25651
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.3166887760162354,
+      "learning_rate": 4.322178165477135e-05,
+      "loss": 1.3466,
+      "step": 25652
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.843368649482727,
+      "learning_rate": 4.322151731430082e-05,
+      "loss": 1.8622,
+      "step": 25653
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0381762981414795,
+      "learning_rate": 4.32212529738303e-05,
+      "loss": 1.3019,
+      "step": 25654
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0950801372528076,
+      "learning_rate": 4.322098863335977e-05,
+      "loss": 1.3437,
+      "step": 25655
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.964647889137268,
+      "learning_rate": 4.3220724292889245e-05,
+      "loss": 1.3065,
+      "step": 25656
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8166953325271606,
+      "learning_rate": 4.322045995241872e-05,
+      "loss": 1.1745,
+      "step": 25657
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0198428630828857,
+      "learning_rate": 4.322019561194819e-05,
+      "loss": 1.2586,
+      "step": 25658
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.79215407371521,
+      "learning_rate": 4.3219931271477665e-05,
+      "loss": 1.2665,
+      "step": 25659
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8579810857772827,
+      "learning_rate": 4.321966693100714e-05,
+      "loss": 1.3555,
+      "step": 25660
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.024364471435547,
+      "learning_rate": 4.321940259053661e-05,
+      "loss": 1.2805,
+      "step": 25661
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7959709167480469,
+      "learning_rate": 4.3219138250066086e-05,
+      "loss": 1.2401,
+      "step": 25662
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2425436973571777,
+      "learning_rate": 4.321887390959556e-05,
+      "loss": 1.2541,
+      "step": 25663
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.170785427093506,
+      "learning_rate": 4.321860956912503e-05,
+      "loss": 1.4102,
+      "step": 25664
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0994484424591064,
+      "learning_rate": 4.3218345228654514e-05,
+      "loss": 1.3212,
+      "step": 25665
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1071014404296875,
+      "learning_rate": 4.321808088818398e-05,
+      "loss": 1.383,
+      "step": 25666
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.3832502365112305,
+      "learning_rate": 4.321781654771346e-05,
+      "loss": 1.4039,
+      "step": 25667
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.3752217292785645,
+      "learning_rate": 4.321755220724293e-05,
+      "loss": 1.5482,
+      "step": 25668
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9244513511657715,
+      "learning_rate": 4.321728786677241e-05,
+      "loss": 1.1117,
+      "step": 25669
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.112675428390503,
+      "learning_rate": 4.321702352630188e-05,
+      "loss": 1.2675,
+      "step": 25670
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8307636976242065,
+      "learning_rate": 4.321675918583135e-05,
+      "loss": 1.0531,
+      "step": 25671
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.5435872077941895,
+      "learning_rate": 4.321649484536083e-05,
+      "loss": 1.1283,
+      "step": 25672
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2123663425445557,
+      "learning_rate": 4.3216230504890295e-05,
+      "loss": 1.387,
+      "step": 25673
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9729793071746826,
+      "learning_rate": 4.3215966164419776e-05,
+      "loss": 1.2787,
+      "step": 25674
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.4019439220428467,
+      "learning_rate": 4.321570182394925e-05,
+      "loss": 1.3333,
+      "step": 25675
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.3092126846313477,
+      "learning_rate": 4.321543748347872e-05,
+      "loss": 1.2252,
+      "step": 25676
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.781851053237915,
+      "learning_rate": 4.3215173143008196e-05,
+      "loss": 1.2967,
+      "step": 25677
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1629772186279297,
+      "learning_rate": 4.321490880253767e-05,
+      "loss": 1.3892,
+      "step": 25678
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.1625847816467285,
+      "learning_rate": 4.3214644462067144e-05,
+      "loss": 1.3322,
+      "step": 25679
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.368826389312744,
+      "learning_rate": 4.321438012159662e-05,
+      "loss": 1.4602,
+      "step": 25680
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7758861780166626,
+      "learning_rate": 4.321411578112609e-05,
+      "loss": 2.2806,
+      "step": 25681
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.235217332839966,
+      "learning_rate": 4.3213851440655564e-05,
+      "loss": 1.3787,
+      "step": 25682
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9900749921798706,
+      "learning_rate": 4.321358710018504e-05,
+      "loss": 1.3583,
+      "step": 25683
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.420419454574585,
+      "learning_rate": 4.321332275971451e-05,
+      "loss": 1.3037,
+      "step": 25684
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.4205427169799805,
+      "learning_rate": 4.321305841924399e-05,
+      "loss": 1.1753,
+      "step": 25685
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.968612551689148,
+      "learning_rate": 4.321279407877346e-05,
+      "loss": 1.1937,
+      "step": 25686
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9781451225280762,
+      "learning_rate": 4.321252973830294e-05,
+      "loss": 1.2732,
+      "step": 25687
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.518091917037964,
+      "learning_rate": 4.3212265397832406e-05,
+      "loss": 1.4058,
+      "step": 25688
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9250911474227905,
+      "learning_rate": 4.3212001057361886e-05,
+      "loss": 1.313,
+      "step": 25689
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.359264612197876,
+      "learning_rate": 4.321173671689136e-05,
+      "loss": 1.3893,
+      "step": 25690
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.278106927871704,
+      "learning_rate": 4.321147237642083e-05,
+      "loss": 1.4476,
+      "step": 25691
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2375879287719727,
+      "learning_rate": 4.321120803595031e-05,
+      "loss": 1.4008,
+      "step": 25692
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2375521659851074,
+      "learning_rate": 4.321094369547978e-05,
+      "loss": 1.271,
+      "step": 25693
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.143124580383301,
+      "learning_rate": 4.3210679355009254e-05,
+      "loss": 1.3136,
+      "step": 25694
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9654511213302612,
+      "learning_rate": 4.321041501453873e-05,
+      "loss": 1.4664,
+      "step": 25695
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.367110013961792,
+      "learning_rate": 4.32101506740682e-05,
+      "loss": 1.3315,
+      "step": 25696
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.727922797203064,
+      "learning_rate": 4.3209886333597675e-05,
+      "loss": 1.2994,
+      "step": 25697
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.6906112432479858,
+      "learning_rate": 4.320962199312715e-05,
+      "loss": 1.1881,
+      "step": 25698
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.6525373458862305,
+      "learning_rate": 4.320935765265662e-05,
+      "loss": 1.2602,
+      "step": 25699
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.591886520385742,
+      "learning_rate": 4.32090933121861e-05,
+      "loss": 1.3849,
+      "step": 25700
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.1857926845550537,
+      "learning_rate": 4.320882897171557e-05,
+      "loss": 1.5206,
+      "step": 25701
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.568416714668274,
+      "learning_rate": 4.320856463124505e-05,
+      "loss": 1.2345,
+      "step": 25702
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.47400164604187,
+      "learning_rate": 4.3208300290774516e-05,
+      "loss": 1.4426,
+      "step": 25703
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2231616973876953,
+      "learning_rate": 4.3208035950303996e-05,
+      "loss": 1.3027,
+      "step": 25704
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.445465564727783,
+      "learning_rate": 4.320777160983347e-05,
+      "loss": 1.4885,
+      "step": 25705
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7208501100540161,
+      "learning_rate": 4.320750726936294e-05,
+      "loss": 1.4813,
+      "step": 25706
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8414722681045532,
+      "learning_rate": 4.320724292889242e-05,
+      "loss": 1.1719,
+      "step": 25707
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.210702896118164,
+      "learning_rate": 4.3206978588421884e-05,
+      "loss": 1.3745,
+      "step": 25708
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9323939085006714,
+      "learning_rate": 4.3206714247951364e-05,
+      "loss": 1.2719,
+      "step": 25709
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.5216100215911865,
+      "learning_rate": 4.320644990748084e-05,
+      "loss": 1.3756,
+      "step": 25710
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.0550575256347656,
+      "learning_rate": 4.320618556701031e-05,
+      "loss": 1.4095,
+      "step": 25711
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.912514567375183,
+      "learning_rate": 4.3205921226539785e-05,
+      "loss": 1.3327,
+      "step": 25712
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1599276065826416,
+      "learning_rate": 4.320565688606926e-05,
+      "loss": 1.4148,
+      "step": 25713
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.75199556350708,
+      "learning_rate": 4.320539254559873e-05,
+      "loss": 1.3795,
+      "step": 25714
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.463776111602783,
+      "learning_rate": 4.320512820512821e-05,
+      "loss": 1.3574,
+      "step": 25715
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.640265703201294,
+      "learning_rate": 4.320486386465768e-05,
+      "loss": 1.2417,
+      "step": 25716
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.056946039199829,
+      "learning_rate": 4.320459952418715e-05,
+      "loss": 1.3116,
+      "step": 25717
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8412514925003052,
+      "learning_rate": 4.3204335183716626e-05,
+      "loss": 1.2911,
+      "step": 25718
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.949272632598877,
+      "learning_rate": 4.32040708432461e-05,
+      "loss": 1.1682,
+      "step": 25719
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2672767639160156,
+      "learning_rate": 4.320380650277558e-05,
+      "loss": 1.296,
+      "step": 25720
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8566642999649048,
+      "learning_rate": 4.320354216230505e-05,
+      "loss": 1.3044,
+      "step": 25721
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.826696753501892,
+      "learning_rate": 4.320327782183453e-05,
+      "loss": 1.3002,
+      "step": 25722
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7166528701782227,
+      "learning_rate": 4.3203013481363994e-05,
+      "loss": 1.3179,
+      "step": 25723
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8478010892868042,
+      "learning_rate": 4.3202749140893475e-05,
+      "loss": 1.3554,
+      "step": 25724
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1173582077026367,
+      "learning_rate": 4.320248480042295e-05,
+      "loss": 1.3637,
+      "step": 25725
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1978297233581543,
+      "learning_rate": 4.320222045995242e-05,
+      "loss": 1.2587,
+      "step": 25726
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0918467044830322,
+      "learning_rate": 4.3201956119481895e-05,
+      "loss": 1.3536,
+      "step": 25727
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.543030023574829,
+      "learning_rate": 4.320169177901137e-05,
+      "loss": 1.4219,
+      "step": 25728
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2535483837127686,
+      "learning_rate": 4.320142743854084e-05,
+      "loss": 1.2056,
+      "step": 25729
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.283493995666504,
+      "learning_rate": 4.3201163098070316e-05,
+      "loss": 1.2187,
+      "step": 25730
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1376993656158447,
+      "learning_rate": 4.320089875759979e-05,
+      "loss": 1.3927,
+      "step": 25731
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9971041679382324,
+      "learning_rate": 4.320063441712926e-05,
+      "loss": 1.4145,
+      "step": 25732
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.682090163230896,
+      "learning_rate": 4.320037007665874e-05,
+      "loss": 1.3745,
+      "step": 25733
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.568704605102539,
+      "learning_rate": 4.320010573618821e-05,
+      "loss": 1.3715,
+      "step": 25734
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0033047199249268,
+      "learning_rate": 4.319984139571769e-05,
+      "loss": 1.3426,
+      "step": 25735
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1992807388305664,
+      "learning_rate": 4.319957705524716e-05,
+      "loss": 1.3246,
+      "step": 25736
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.118034839630127,
+      "learning_rate": 4.319931271477664e-05,
+      "loss": 1.4047,
+      "step": 25737
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.789036989212036,
+      "learning_rate": 4.3199048374306105e-05,
+      "loss": 1.212,
+      "step": 25738
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.512073040008545,
+      "learning_rate": 4.3198784033835585e-05,
+      "loss": 1.2041,
+      "step": 25739
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.6846489906311035,
+      "learning_rate": 4.319851969336506e-05,
+      "loss": 1.4082,
+      "step": 25740
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1728262901306152,
+      "learning_rate": 4.319825535289453e-05,
+      "loss": 1.2726,
+      "step": 25741
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8271220922470093,
+      "learning_rate": 4.3197991012424006e-05,
+      "loss": 1.1575,
+      "step": 25742
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.5010995864868164,
+      "learning_rate": 4.319772667195347e-05,
+      "loss": 1.4727,
+      "step": 25743
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2634470462799072,
+      "learning_rate": 4.319746233148295e-05,
+      "loss": 1.3637,
+      "step": 25744
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.050382137298584,
+      "learning_rate": 4.3197197991012426e-05,
+      "loss": 1.2763,
+      "step": 25745
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9554964303970337,
+      "learning_rate": 4.31969336505419e-05,
+      "loss": 1.2981,
+      "step": 25746
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.291938066482544,
+      "learning_rate": 4.3196669310071374e-05,
+      "loss": 1.4021,
+      "step": 25747
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.1207118034362793,
+      "learning_rate": 4.319640496960085e-05,
+      "loss": 1.2535,
+      "step": 25748
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8158502578735352,
+      "learning_rate": 4.319614062913032e-05,
+      "loss": 1.2444,
+      "step": 25749
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2115187644958496,
+      "learning_rate": 4.31958762886598e-05,
+      "loss": 1.4048,
+      "step": 25750
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0885393619537354,
+      "learning_rate": 4.319561194818927e-05,
+      "loss": 1.3158,
+      "step": 25751
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8555591106414795,
+      "learning_rate": 4.319534760771874e-05,
+      "loss": 1.3146,
+      "step": 25752
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.982746958732605,
+      "learning_rate": 4.3195083267248215e-05,
+      "loss": 1.4366,
+      "step": 25753
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.992343544960022,
+      "learning_rate": 4.319481892677769e-05,
+      "loss": 1.2701,
+      "step": 25754
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8248478174209595,
+      "learning_rate": 4.319455458630717e-05,
+      "loss": 1.2553,
+      "step": 25755
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.3919203281402588,
+      "learning_rate": 4.3194290245836636e-05,
+      "loss": 1.0131,
+      "step": 25756
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7821837663650513,
+      "learning_rate": 4.3194025905366116e-05,
+      "loss": 1.3592,
+      "step": 25757
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.41534686088562,
+      "learning_rate": 4.319376156489558e-05,
+      "loss": 1.3636,
+      "step": 25758
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0167880058288574,
+      "learning_rate": 4.319349722442506e-05,
+      "loss": 1.3025,
+      "step": 25759
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.6081129312515259,
+      "learning_rate": 4.319323288395454e-05,
+      "loss": 2.0652,
+      "step": 25760
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7235556840896606,
+      "learning_rate": 4.319296854348401e-05,
+      "loss": 1.2942,
+      "step": 25761
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7403758764266968,
+      "learning_rate": 4.3192704203013484e-05,
+      "loss": 1.1343,
+      "step": 25762
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.502094268798828,
+      "learning_rate": 4.319243986254296e-05,
+      "loss": 1.5235,
+      "step": 25763
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.218811273574829,
+      "learning_rate": 4.319217552207243e-05,
+      "loss": 1.1122,
+      "step": 25764
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7798250913619995,
+      "learning_rate": 4.3191911181601905e-05,
+      "loss": 1.2172,
+      "step": 25765
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8260562419891357,
+      "learning_rate": 4.319164684113138e-05,
+      "loss": 1.2113,
+      "step": 25766
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.589585304260254,
+      "learning_rate": 4.319138250066085e-05,
+      "loss": 1.3463,
+      "step": 25767
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.323979616165161,
+      "learning_rate": 4.3191118160190325e-05,
+      "loss": 1.4134,
+      "step": 25768
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.4369544982910156,
+      "learning_rate": 4.31908538197198e-05,
+      "loss": 1.309,
+      "step": 25769
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.067528009414673,
+      "learning_rate": 4.319058947924928e-05,
+      "loss": 1.3565,
+      "step": 25770
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9945038557052612,
+      "learning_rate": 4.3190325138778746e-05,
+      "loss": 1.2401,
+      "step": 25771
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9196876287460327,
+      "learning_rate": 4.3190060798308226e-05,
+      "loss": 1.2078,
+      "step": 25772
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7424485683441162,
+      "learning_rate": 4.318979645783769e-05,
+      "loss": 1.354,
+      "step": 25773
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.067445993423462,
+      "learning_rate": 4.3189532117367173e-05,
+      "loss": 1.3025,
+      "step": 25774
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8577193021774292,
+      "learning_rate": 4.318926777689665e-05,
+      "loss": 1.2342,
+      "step": 25775
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9900633096694946,
+      "learning_rate": 4.318900343642612e-05,
+      "loss": 1.3179,
+      "step": 25776
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0540153980255127,
+      "learning_rate": 4.3188739095955594e-05,
+      "loss": 1.3851,
+      "step": 25777
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.6176512241363525,
+      "learning_rate": 4.318847475548507e-05,
+      "loss": 1.1932,
+      "step": 25778
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.5193393230438232,
+      "learning_rate": 4.318821041501454e-05,
+      "loss": 1.1058,
+      "step": 25779
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.052753210067749,
+      "learning_rate": 4.3187946074544015e-05,
+      "loss": 1.1716,
+      "step": 25780
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.767323613166809,
+      "learning_rate": 4.318768173407349e-05,
+      "loss": 1.1914,
+      "step": 25781
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.4154510498046875,
+      "learning_rate": 4.318741739360296e-05,
+      "loss": 1.3041,
+      "step": 25782
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0705223083496094,
+      "learning_rate": 4.3187153053132436e-05,
+      "loss": 1.46,
+      "step": 25783
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9564905166625977,
+      "learning_rate": 4.318688871266191e-05,
+      "loss": 1.239,
+      "step": 25784
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8352235555648804,
+      "learning_rate": 4.318662437219139e-05,
+      "loss": 1.2578,
+      "step": 25785
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7216911315917969,
+      "learning_rate": 4.3186360031720856e-05,
+      "loss": 2.2397,
+      "step": 25786
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9545243978500366,
+      "learning_rate": 4.318609569125034e-05,
+      "loss": 1.24,
+      "step": 25787
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7528927326202393,
+      "learning_rate": 4.3185831350779803e-05,
+      "loss": 1.1938,
+      "step": 25788
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.200640916824341,
+      "learning_rate": 4.318556701030928e-05,
+      "loss": 1.3934,
+      "step": 25789
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9725655317306519,
+      "learning_rate": 4.318530266983876e-05,
+      "loss": 1.3083,
+      "step": 25790
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8387486934661865,
+      "learning_rate": 4.3185038329368224e-05,
+      "loss": 1.213,
+      "step": 25791
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2772436141967773,
+      "learning_rate": 4.3184773988897705e-05,
+      "loss": 1.4299,
+      "step": 25792
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.549783706665039,
+      "learning_rate": 4.318450964842717e-05,
+      "loss": 1.1271,
+      "step": 25793
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.238133192062378,
+      "learning_rate": 4.318424530795665e-05,
+      "loss": 1.4056,
+      "step": 25794
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.2164108753204346,
+      "learning_rate": 4.3183980967486125e-05,
+      "loss": 1.3244,
+      "step": 25795
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.5937111377716064,
+      "learning_rate": 4.31837166270156e-05,
+      "loss": 1.248,
+      "step": 25796
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.808320164680481,
+      "learning_rate": 4.318345228654507e-05,
+      "loss": 1.1251,
+      "step": 25797
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9633876085281372,
+      "learning_rate": 4.3183187946074546e-05,
+      "loss": 1.2342,
+      "step": 25798
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0720431804656982,
+      "learning_rate": 4.318292360560402e-05,
+      "loss": 1.2616,
+      "step": 25799
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.47359561920166,
+      "learning_rate": 4.318265926513349e-05,
+      "loss": 1.2406,
+      "step": 25800
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7111186981201172,
+      "learning_rate": 4.318239492466297e-05,
+      "loss": 1.2317,
+      "step": 25801
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.4883334636688232,
+      "learning_rate": 4.318213058419244e-05,
+      "loss": 1.3485,
+      "step": 25802
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.4495248794555664,
+      "learning_rate": 4.3181866243721914e-05,
+      "loss": 1.4133,
+      "step": 25803
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.940373420715332,
+      "learning_rate": 4.318160190325139e-05,
+      "loss": 1.267,
+      "step": 25804
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9131419658660889,
+      "learning_rate": 4.318133756278087e-05,
+      "loss": 1.2041,
+      "step": 25805
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.179844379425049,
+      "learning_rate": 4.3181073222310335e-05,
+      "loss": 1.3305,
+      "step": 25806
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9998230934143066,
+      "learning_rate": 4.3180808881839815e-05,
+      "loss": 1.4431,
+      "step": 25807
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.0348260402679443,
+      "learning_rate": 4.318054454136928e-05,
+      "loss": 1.2933,
+      "step": 25808
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9092628955841064,
+      "learning_rate": 4.318028020089876e-05,
+      "loss": 1.2742,
+      "step": 25809
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9727572202682495,
+      "learning_rate": 4.3180015860428236e-05,
+      "loss": 1.2762,
+      "step": 25810
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.972747802734375,
+      "learning_rate": 4.317975151995771e-05,
+      "loss": 1.3004,
+      "step": 25811
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9198651313781738,
+      "learning_rate": 4.317948717948718e-05,
+      "loss": 1.198,
+      "step": 25812
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.6780662536621094,
+      "learning_rate": 4.3179222839016656e-05,
+      "loss": 1.1705,
+      "step": 25813
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.1857106685638428,
+      "learning_rate": 4.317895849854613e-05,
+      "loss": 1.2502,
+      "step": 25814
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8220183849334717,
+      "learning_rate": 4.3178694158075603e-05,
+      "loss": 1.2737,
+      "step": 25815
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9022029638290405,
+      "learning_rate": 4.317842981760508e-05,
+      "loss": 1.2328,
+      "step": 25816
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.943379521369934,
+      "learning_rate": 4.317816547713455e-05,
+      "loss": 1.324,
+      "step": 25817
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.4639203548431396,
+      "learning_rate": 4.3177901136664024e-05,
+      "loss": 1.3052,
+      "step": 25818
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.470226287841797,
+      "learning_rate": 4.31776367961935e-05,
+      "loss": 1.4352,
+      "step": 25819
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.220491886138916,
+      "learning_rate": 4.317737245572298e-05,
+      "loss": 1.269,
+      "step": 25820
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7318339347839355,
+      "learning_rate": 4.3177108115252445e-05,
+      "loss": 1.3011,
+      "step": 25821
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3003082275390625,
+      "learning_rate": 4.3176843774781925e-05,
+      "loss": 1.3144,
+      "step": 25822
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8001916408538818,
+      "learning_rate": 4.317657943431139e-05,
+      "loss": 1.2362,
+      "step": 25823
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9482420682907104,
+      "learning_rate": 4.3176315093840866e-05,
+      "loss": 1.1017,
+      "step": 25824
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0333473682403564,
+      "learning_rate": 4.3176050753370346e-05,
+      "loss": 1.241,
+      "step": 25825
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.8507421016693115,
+      "learning_rate": 4.317578641289981e-05,
+      "loss": 1.4664,
+      "step": 25826
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.118898630142212,
+      "learning_rate": 4.317552207242929e-05,
+      "loss": 1.389,
+      "step": 25827
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1351301670074463,
+      "learning_rate": 4.317525773195876e-05,
+      "loss": 1.4431,
+      "step": 25828
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8943370580673218,
+      "learning_rate": 4.317499339148824e-05,
+      "loss": 1.1974,
+      "step": 25829
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9444841146469116,
+      "learning_rate": 4.3174729051017714e-05,
+      "loss": 1.2337,
+      "step": 25830
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.786808490753174,
+      "learning_rate": 4.317446471054719e-05,
+      "loss": 1.3406,
+      "step": 25831
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.5157630443573,
+      "learning_rate": 4.317420037007666e-05,
+      "loss": 1.3599,
+      "step": 25832
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0164973735809326,
+      "learning_rate": 4.3173936029606134e-05,
+      "loss": 1.3262,
+      "step": 25833
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.637988805770874,
+      "learning_rate": 4.317367168913561e-05,
+      "loss": 1.3045,
+      "step": 25834
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.533876419067383,
+      "learning_rate": 4.317340734866508e-05,
+      "loss": 1.7473,
+      "step": 25835
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0882115364074707,
+      "learning_rate": 4.3173143008194555e-05,
+      "loss": 1.3593,
+      "step": 25836
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7570972442626953,
+      "learning_rate": 4.317287866772403e-05,
+      "loss": 1.2282,
+      "step": 25837
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9374910593032837,
+      "learning_rate": 4.31726143272535e-05,
+      "loss": 1.3479,
+      "step": 25838
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.823009967803955,
+      "learning_rate": 4.3172349986782976e-05,
+      "loss": 1.3349,
+      "step": 25839
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0422613620758057,
+      "learning_rate": 4.3172085646312456e-05,
+      "loss": 1.5373,
+      "step": 25840
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8578404188156128,
+      "learning_rate": 4.317182130584192e-05,
+      "loss": 1.2939,
+      "step": 25841
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9975411891937256,
+      "learning_rate": 4.3171556965371403e-05,
+      "loss": 1.262,
+      "step": 25842
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.264584541320801,
+      "learning_rate": 4.317129262490087e-05,
+      "loss": 1.207,
+      "step": 25843
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.065560817718506,
+      "learning_rate": 4.317102828443035e-05,
+      "loss": 1.3023,
+      "step": 25844
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.5634419918060303,
+      "learning_rate": 4.3170763943959824e-05,
+      "loss": 1.2682,
+      "step": 25845
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3608105182647705,
+      "learning_rate": 4.31704996034893e-05,
+      "loss": 1.3955,
+      "step": 25846
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8163251876831055,
+      "learning_rate": 4.317023526301877e-05,
+      "loss": 1.2622,
+      "step": 25847
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7879959344863892,
+      "learning_rate": 4.3169970922548245e-05,
+      "loss": 1.211,
+      "step": 25848
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.5830774307250977,
+      "learning_rate": 4.316970658207772e-05,
+      "loss": 1.2513,
+      "step": 25849
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0468785762786865,
+      "learning_rate": 4.316944224160719e-05,
+      "loss": 1.27,
+      "step": 25850
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6793012619018555,
+      "learning_rate": 4.3169177901136666e-05,
+      "loss": 1.1282,
+      "step": 25851
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.035647392272949,
+      "learning_rate": 4.316891356066614e-05,
+      "loss": 1.4507,
+      "step": 25852
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6997966766357422,
+      "learning_rate": 4.316864922019561e-05,
+      "loss": 1.344,
+      "step": 25853
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.045243978500366,
+      "learning_rate": 4.3168384879725086e-05,
+      "loss": 1.1559,
+      "step": 25854
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0231800079345703,
+      "learning_rate": 4.3168120539254567e-05,
+      "loss": 1.2978,
+      "step": 25855
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.807012677192688,
+      "learning_rate": 4.3167856198784033e-05,
+      "loss": 1.1731,
+      "step": 25856
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7799758911132812,
+      "learning_rate": 4.3167591858313514e-05,
+      "loss": 1.2852,
+      "step": 25857
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8551056385040283,
+      "learning_rate": 4.316732751784298e-05,
+      "loss": 1.276,
+      "step": 25858
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1634109020233154,
+      "learning_rate": 4.316706317737246e-05,
+      "loss": 1.3684,
+      "step": 25859
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0743963718414307,
+      "learning_rate": 4.3166798836901934e-05,
+      "loss": 1.2851,
+      "step": 25860
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3965139389038086,
+      "learning_rate": 4.31665344964314e-05,
+      "loss": 1.2519,
+      "step": 25861
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0546481609344482,
+      "learning_rate": 4.316627015596088e-05,
+      "loss": 1.2403,
+      "step": 25862
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7575770616531372,
+      "learning_rate": 4.316600581549035e-05,
+      "loss": 1.2595,
+      "step": 25863
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1507010459899902,
+      "learning_rate": 4.316574147501983e-05,
+      "loss": 1.2455,
+      "step": 25864
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7726346254348755,
+      "learning_rate": 4.31654771345493e-05,
+      "loss": 1.1247,
+      "step": 25865
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3219330310821533,
+      "learning_rate": 4.3165212794078776e-05,
+      "loss": 1.3786,
+      "step": 25866
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7878413200378418,
+      "learning_rate": 4.316494845360825e-05,
+      "loss": 1.1965,
+      "step": 25867
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.5263428688049316,
+      "learning_rate": 4.316468411313772e-05,
+      "loss": 1.4975,
+      "step": 25868
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.5115439891815186,
+      "learning_rate": 4.3164419772667197e-05,
+      "loss": 1.2927,
+      "step": 25869
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7903802394866943,
+      "learning_rate": 4.316415543219667e-05,
+      "loss": 1.9461,
+      "step": 25870
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9946480989456177,
+      "learning_rate": 4.3163891091726144e-05,
+      "loss": 1.2304,
+      "step": 25871
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.817501187324524,
+      "learning_rate": 4.316362675125562e-05,
+      "loss": 1.2033,
+      "step": 25872
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7154475450515747,
+      "learning_rate": 4.316336241078509e-05,
+      "loss": 1.2507,
+      "step": 25873
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8181118965148926,
+      "learning_rate": 4.3163098070314564e-05,
+      "loss": 1.32,
+      "step": 25874
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9765546321868896,
+      "learning_rate": 4.3162833729844045e-05,
+      "loss": 1.2974,
+      "step": 25875
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8143854141235352,
+      "learning_rate": 4.316256938937351e-05,
+      "loss": 1.2432,
+      "step": 25876
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.67919921875,
+      "learning_rate": 4.316230504890299e-05,
+      "loss": 1.1698,
+      "step": 25877
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7719498872756958,
+      "learning_rate": 4.316204070843246e-05,
+      "loss": 1.6183,
+      "step": 25878
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6889439821243286,
+      "learning_rate": 4.316177636796194e-05,
+      "loss": 1.2494,
+      "step": 25879
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.5759174823760986,
+      "learning_rate": 4.316151202749141e-05,
+      "loss": 1.5184,
+      "step": 25880
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8921129703521729,
+      "learning_rate": 4.3161247687020886e-05,
+      "loss": 1.2987,
+      "step": 25881
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.4645214080810547,
+      "learning_rate": 4.316098334655036e-05,
+      "loss": 1.3081,
+      "step": 25882
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.882934808731079,
+      "learning_rate": 4.316071900607983e-05,
+      "loss": 1.3808,
+      "step": 25883
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8666495084762573,
+      "learning_rate": 4.316045466560931e-05,
+      "loss": 1.404,
+      "step": 25884
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0616965293884277,
+      "learning_rate": 4.316019032513878e-05,
+      "loss": 1.4386,
+      "step": 25885
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1573293209075928,
+      "learning_rate": 4.3159925984668254e-05,
+      "loss": 1.4794,
+      "step": 25886
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9150712490081787,
+      "learning_rate": 4.315966164419773e-05,
+      "loss": 1.1728,
+      "step": 25887
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0365617275238037,
+      "learning_rate": 4.31593973037272e-05,
+      "loss": 1.4677,
+      "step": 25888
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9500457048416138,
+      "learning_rate": 4.3159132963256675e-05,
+      "loss": 1.2623,
+      "step": 25889
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8948472738265991,
+      "learning_rate": 4.3158868622786155e-05,
+      "loss": 1.2903,
+      "step": 25890
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.295740842819214,
+      "learning_rate": 4.315860428231562e-05,
+      "loss": 1.1823,
+      "step": 25891
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.85014009475708,
+      "learning_rate": 4.31583399418451e-05,
+      "loss": 1.2119,
+      "step": 25892
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.4141488075256348,
+      "learning_rate": 4.315807560137457e-05,
+      "loss": 1.5471,
+      "step": 25893
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6719341278076172,
+      "learning_rate": 4.315781126090405e-05,
+      "loss": 1.1385,
+      "step": 25894
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0507569313049316,
+      "learning_rate": 4.315754692043352e-05,
+      "loss": 1.5474,
+      "step": 25895
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6937047243118286,
+      "learning_rate": 4.3157282579962997e-05,
+      "loss": 1.1855,
+      "step": 25896
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9628610610961914,
+      "learning_rate": 4.315701823949247e-05,
+      "loss": 1.289,
+      "step": 25897
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8858973979949951,
+      "learning_rate": 4.315675389902194e-05,
+      "loss": 1.2947,
+      "step": 25898
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8666797876358032,
+      "learning_rate": 4.315648955855142e-05,
+      "loss": 1.3593,
+      "step": 25899
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9206998348236084,
+      "learning_rate": 4.315622521808089e-05,
+      "loss": 1.2336,
+      "step": 25900
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8441894054412842,
+      "learning_rate": 4.3155960877610364e-05,
+      "loss": 1.3277,
+      "step": 25901
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1394810676574707,
+      "learning_rate": 4.315569653713984e-05,
+      "loss": 1.38,
+      "step": 25902
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0636041164398193,
+      "learning_rate": 4.315543219666931e-05,
+      "loss": 1.3136,
+      "step": 25903
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.323302745819092,
+      "learning_rate": 4.3155167856198785e-05,
+      "loss": 1.2465,
+      "step": 25904
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0348031520843506,
+      "learning_rate": 4.3154903515728265e-05,
+      "loss": 1.3297,
+      "step": 25905
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.810878038406372,
+      "learning_rate": 4.315463917525773e-05,
+      "loss": 1.1366,
+      "step": 25906
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.531215190887451,
+      "learning_rate": 4.3154374834787206e-05,
+      "loss": 1.3259,
+      "step": 25907
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.589050531387329,
+      "learning_rate": 4.315411049431668e-05,
+      "loss": 1.202,
+      "step": 25908
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8596631288528442,
+      "learning_rate": 4.315384615384615e-05,
+      "loss": 1.4084,
+      "step": 25909
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.2814157009124756,
+      "learning_rate": 4.315358181337563e-05,
+      "loss": 1.3049,
+      "step": 25910
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9841506481170654,
+      "learning_rate": 4.31533174729051e-05,
+      "loss": 1.324,
+      "step": 25911
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.6819753646850586,
+      "learning_rate": 4.315305313243458e-05,
+      "loss": 1.3249,
+      "step": 25912
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.2291646003723145,
+      "learning_rate": 4.315278879196405e-05,
+      "loss": 1.4914,
+      "step": 25913
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0454258918762207,
+      "learning_rate": 4.315252445149353e-05,
+      "loss": 1.2273,
+      "step": 25914
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.2968950271606445,
+      "learning_rate": 4.3152260111023e-05,
+      "loss": 1.5395,
+      "step": 25915
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.742529034614563,
+      "learning_rate": 4.3151995770552475e-05,
+      "loss": 1.1406,
+      "step": 25916
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.559053659439087,
+      "learning_rate": 4.315173143008195e-05,
+      "loss": 1.6961,
+      "step": 25917
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.5822793245315552,
+      "learning_rate": 4.315146708961142e-05,
+      "loss": 1.2244,
+      "step": 25918
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.246023416519165,
+      "learning_rate": 4.3151202749140895e-05,
+      "loss": 1.5337,
+      "step": 25919
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.089249610900879,
+      "learning_rate": 4.315093840867037e-05,
+      "loss": 1.3601,
+      "step": 25920
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.293973684310913,
+      "learning_rate": 4.315067406819984e-05,
+      "loss": 1.222,
+      "step": 25921
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9322069883346558,
+      "learning_rate": 4.3150409727729316e-05,
+      "loss": 1.4368,
+      "step": 25922
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7771655321121216,
+      "learning_rate": 4.315014538725879e-05,
+      "loss": 1.3229,
+      "step": 25923
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.752143144607544,
+      "learning_rate": 4.314988104678826e-05,
+      "loss": 1.2527,
+      "step": 25924
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.566758155822754,
+      "learning_rate": 4.3149616706317744e-05,
+      "loss": 1.3785,
+      "step": 25925
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.080380916595459,
+      "learning_rate": 4.314935236584721e-05,
+      "loss": 1.3652,
+      "step": 25926
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3332407474517822,
+      "learning_rate": 4.314908802537669e-05,
+      "loss": 1.6417,
+      "step": 25927
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.165142774581909,
+      "learning_rate": 4.314882368490616e-05,
+      "loss": 1.3604,
+      "step": 25928
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8657331466674805,
+      "learning_rate": 4.314855934443564e-05,
+      "loss": 1.2551,
+      "step": 25929
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.121293067932129,
+      "learning_rate": 4.314829500396511e-05,
+      "loss": 1.3446,
+      "step": 25930
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3049402236938477,
+      "learning_rate": 4.3148030663494585e-05,
+      "loss": 1.4846,
+      "step": 25931
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8512316942214966,
+      "learning_rate": 4.314776632302406e-05,
+      "loss": 1.2855,
+      "step": 25932
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.065711498260498,
+      "learning_rate": 4.3147501982553525e-05,
+      "loss": 1.3851,
+      "step": 25933
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.559537172317505,
+      "learning_rate": 4.3147237642083006e-05,
+      "loss": 1.4048,
+      "step": 25934
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9361693859100342,
+      "learning_rate": 4.314697330161248e-05,
+      "loss": 1.3503,
+      "step": 25935
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.98709774017334,
+      "learning_rate": 4.314670896114195e-05,
+      "loss": 1.3357,
+      "step": 25936
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0006937980651855,
+      "learning_rate": 4.3146444620671427e-05,
+      "loss": 1.2575,
+      "step": 25937
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.417933225631714,
+      "learning_rate": 4.31461802802009e-05,
+      "loss": 1.3975,
+      "step": 25938
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.8132476806640625,
+      "learning_rate": 4.3145915939730374e-05,
+      "loss": 1.4141,
+      "step": 25939
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0118637084960938,
+      "learning_rate": 4.3145651599259854e-05,
+      "loss": 1.4073,
+      "step": 25940
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8684415817260742,
+      "learning_rate": 4.314538725878932e-05,
+      "loss": 1.1919,
+      "step": 25941
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.85011625289917,
+      "learning_rate": 4.3145122918318794e-05,
+      "loss": 1.2185,
+      "step": 25942
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.946128010749817,
+      "learning_rate": 4.314485857784827e-05,
+      "loss": 1.269,
+      "step": 25943
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8071625232696533,
+      "learning_rate": 4.314459423737774e-05,
+      "loss": 1.2756,
+      "step": 25944
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.025888442993164,
+      "learning_rate": 4.314432989690722e-05,
+      "loss": 1.3356,
+      "step": 25945
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6863393783569336,
+      "learning_rate": 4.314406555643669e-05,
+      "loss": 1.3492,
+      "step": 25946
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.432506799697876,
+      "learning_rate": 4.314380121596617e-05,
+      "loss": 1.4654,
+      "step": 25947
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3569788932800293,
+      "learning_rate": 4.3143536875495636e-05,
+      "loss": 1.2866,
+      "step": 25948
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.914459228515625,
+      "learning_rate": 4.3143272535025116e-05,
+      "loss": 1.3804,
+      "step": 25949
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6137839555740356,
+      "learning_rate": 4.314300819455459e-05,
+      "loss": 1.1441,
+      "step": 25950
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.146704912185669,
+      "learning_rate": 4.314274385408406e-05,
+      "loss": 1.2344,
+      "step": 25951
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1786136627197266,
+      "learning_rate": 4.314247951361354e-05,
+      "loss": 1.2859,
+      "step": 25952
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.35036563873291,
+      "learning_rate": 4.314221517314301e-05,
+      "loss": 1.3094,
+      "step": 25953
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.316638469696045,
+      "learning_rate": 4.3141950832672484e-05,
+      "loss": 1.2896,
+      "step": 25954
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3126964569091797,
+      "learning_rate": 4.314168649220196e-05,
+      "loss": 1.4008,
+      "step": 25955
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8885546922683716,
+      "learning_rate": 4.314142215173143e-05,
+      "loss": 1.3587,
+      "step": 25956
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.338135004043579,
+      "learning_rate": 4.3141157811260905e-05,
+      "loss": 1.4398,
+      "step": 25957
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1856186389923096,
+      "learning_rate": 4.314089347079038e-05,
+      "loss": 1.1481,
+      "step": 25958
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.4913101196289062,
+      "learning_rate": 4.314062913031985e-05,
+      "loss": 1.296,
+      "step": 25959
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8781687021255493,
+      "learning_rate": 4.314036478984933e-05,
+      "loss": 1.3057,
+      "step": 25960
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9696720838546753,
+      "learning_rate": 4.31401004493788e-05,
+      "loss": 1.4287,
+      "step": 25961
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.5782651901245117,
+      "learning_rate": 4.313983610890828e-05,
+      "loss": 1.3722,
+      "step": 25962
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8775711059570312,
+      "learning_rate": 4.3139571768437746e-05,
+      "loss": 1.3311,
+      "step": 25963
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.087002992630005,
+      "learning_rate": 4.3139307427967226e-05,
+      "loss": 1.2674,
+      "step": 25964
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.655440092086792,
+      "learning_rate": 4.31390430874967e-05,
+      "loss": 1.2594,
+      "step": 25965
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.973076343536377,
+      "learning_rate": 4.3138778747026174e-05,
+      "loss": 1.3361,
+      "step": 25966
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.450269937515259,
+      "learning_rate": 4.313851440655565e-05,
+      "loss": 1.2512,
+      "step": 25967
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.605684518814087,
+      "learning_rate": 4.313825006608512e-05,
+      "loss": 1.6229,
+      "step": 25968
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.091944932937622,
+      "learning_rate": 4.3137985725614594e-05,
+      "loss": 1.3341,
+      "step": 25969
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.284642457962036,
+      "learning_rate": 4.313772138514407e-05,
+      "loss": 1.3672,
+      "step": 25970
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.2102768421173096,
+      "learning_rate": 4.313745704467354e-05,
+      "loss": 1.3856,
+      "step": 25971
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.737747073173523,
+      "learning_rate": 4.3137192704203015e-05,
+      "loss": 1.3064,
+      "step": 25972
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1046953201293945,
+      "learning_rate": 4.313692836373249e-05,
+      "loss": 1.373,
+      "step": 25973
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7517744302749634,
+      "learning_rate": 4.313666402326196e-05,
+      "loss": 1.211,
+      "step": 25974
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1437785625457764,
+      "learning_rate": 4.313639968279144e-05,
+      "loss": 1.587,
+      "step": 25975
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3980369567871094,
+      "learning_rate": 4.313613534232091e-05,
+      "loss": 1.3042,
+      "step": 25976
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.836466908454895,
+      "learning_rate": 4.313587100185039e-05,
+      "loss": 1.302,
+      "step": 25977
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.584108591079712,
+      "learning_rate": 4.3135606661379856e-05,
+      "loss": 1.3006,
+      "step": 25978
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.335113763809204,
+      "learning_rate": 4.313534232090933e-05,
+      "loss": 1.3026,
+      "step": 25979
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.6524624824523926,
+      "learning_rate": 4.313507798043881e-05,
+      "loss": 1.6483,
+      "step": 25980
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.104865074157715,
+      "learning_rate": 4.313481363996828e-05,
+      "loss": 1.3708,
+      "step": 25981
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.4188430309295654,
+      "learning_rate": 4.313454929949776e-05,
+      "loss": 1.3179,
+      "step": 25982
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.304847478866577,
+      "learning_rate": 4.3134284959027224e-05,
+      "loss": 1.3875,
+      "step": 25983
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8196712732315063,
+      "learning_rate": 4.3134020618556705e-05,
+      "loss": 1.238,
+      "step": 25984
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.5957411527633667,
+      "learning_rate": 4.313375627808618e-05,
+      "loss": 1.155,
+      "step": 25985
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.141073703765869,
+      "learning_rate": 4.313349193761565e-05,
+      "loss": 1.5183,
+      "step": 25986
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9721471071243286,
+      "learning_rate": 4.3133227597145125e-05,
+      "loss": 1.3238,
+      "step": 25987
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8800212144851685,
+      "learning_rate": 4.31329632566746e-05,
+      "loss": 1.2905,
+      "step": 25988
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.179550886154175,
+      "learning_rate": 4.313269891620407e-05,
+      "loss": 1.2641,
+      "step": 25989
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6289232969284058,
+      "learning_rate": 4.3132434575733546e-05,
+      "loss": 1.2749,
+      "step": 25990
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8452352285385132,
+      "learning_rate": 4.313217023526302e-05,
+      "loss": 1.2829,
+      "step": 25991
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9879032373428345,
+      "learning_rate": 4.313190589479249e-05,
+      "loss": 1.2407,
+      "step": 25992
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.158116102218628,
+      "learning_rate": 4.313164155432197e-05,
+      "loss": 1.3509,
+      "step": 25993
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.4006423950195312,
+      "learning_rate": 4.313137721385144e-05,
+      "loss": 1.3552,
+      "step": 25994
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9388312101364136,
+      "learning_rate": 4.313111287338092e-05,
+      "loss": 1.3157,
+      "step": 25995
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0597338676452637,
+      "learning_rate": 4.313084853291039e-05,
+      "loss": 1.2974,
+      "step": 25996
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8860396146774292,
+      "learning_rate": 4.313058419243987e-05,
+      "loss": 1.2704,
+      "step": 25997
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.350447177886963,
+      "learning_rate": 4.3130319851969335e-05,
+      "loss": 1.2144,
+      "step": 25998
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1974101066589355,
+      "learning_rate": 4.3130055511498815e-05,
+      "loss": 1.9073,
+      "step": 25999
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1917169094085693,
+      "learning_rate": 4.312979117102829e-05,
+      "loss": 1.2861,
+      "step": 26000
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9981803894042969,
+      "learning_rate": 4.312952683055776e-05,
+      "loss": 1.5297,
+      "step": 26001
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.1777405738830566,
+      "learning_rate": 4.3129262490087236e-05,
+      "loss": 1.4633,
+      "step": 26002
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.9404726028442383,
+      "learning_rate": 4.312899814961671e-05,
+      "loss": 1.5712,
+      "step": 26003
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.346740245819092,
+      "learning_rate": 4.312873380914618e-05,
+      "loss": 1.3265,
+      "step": 26004
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.068554401397705,
+      "learning_rate": 4.3128469468675656e-05,
+      "loss": 1.3232,
+      "step": 26005
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0867953300476074,
+      "learning_rate": 4.312820512820513e-05,
+      "loss": 1.2895,
+      "step": 26006
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7552075386047363,
+      "learning_rate": 4.3127940787734604e-05,
+      "loss": 1.127,
+      "step": 26007
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.880653738975525,
+      "learning_rate": 4.312767644726408e-05,
+      "loss": 1.3018,
+      "step": 26008
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.2821521759033203,
+      "learning_rate": 4.312741210679355e-05,
+      "loss": 1.2422,
+      "step": 26009
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8926494121551514,
+      "learning_rate": 4.312714776632303e-05,
+      "loss": 1.1886,
+      "step": 26010
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.3234758377075195,
+      "learning_rate": 4.31268834258525e-05,
+      "loss": 1.368,
+      "step": 26011
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9832146167755127,
+      "learning_rate": 4.312661908538198e-05,
+      "loss": 1.2231,
+      "step": 26012
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1005077362060547,
+      "learning_rate": 4.3126354744911445e-05,
+      "loss": 1.2672,
+      "step": 26013
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.819298267364502,
+      "learning_rate": 4.312609040444092e-05,
+      "loss": 1.4875,
+      "step": 26014
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1035611629486084,
+      "learning_rate": 4.31258260639704e-05,
+      "loss": 1.3708,
+      "step": 26015
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8284322023391724,
+      "learning_rate": 4.3125561723499866e-05,
+      "loss": 1.1275,
+      "step": 26016
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.7924888134002686,
+      "learning_rate": 4.3125297383029346e-05,
+      "loss": 1.1939,
+      "step": 26017
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7325235605239868,
+      "learning_rate": 4.312503304255881e-05,
+      "loss": 1.1966,
+      "step": 26018
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.623087763786316,
+      "learning_rate": 4.312476870208829e-05,
+      "loss": 1.0223,
+      "step": 26019
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1791985034942627,
+      "learning_rate": 4.312450436161777e-05,
+      "loss": 1.3422,
+      "step": 26020
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9437146186828613,
+      "learning_rate": 4.312424002114724e-05,
+      "loss": 1.3121,
+      "step": 26021
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9067491292953491,
+      "learning_rate": 4.3123975680676714e-05,
+      "loss": 2.5926,
+      "step": 26022
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.460540533065796,
+      "learning_rate": 4.312371134020619e-05,
+      "loss": 1.2789,
+      "step": 26023
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.933708429336548,
+      "learning_rate": 4.312344699973566e-05,
+      "loss": 1.3113,
+      "step": 26024
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8028342723846436,
+      "learning_rate": 4.3123182659265135e-05,
+      "loss": 1.2206,
+      "step": 26025
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.4685921669006348,
+      "learning_rate": 4.312291831879461e-05,
+      "loss": 1.3942,
+      "step": 26026
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.125411033630371,
+      "learning_rate": 4.312265397832408e-05,
+      "loss": 1.3449,
+      "step": 26027
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1571173667907715,
+      "learning_rate": 4.3122389637853555e-05,
+      "loss": 1.2881,
+      "step": 26028
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.13181734085083,
+      "learning_rate": 4.312212529738303e-05,
+      "loss": 1.5162,
+      "step": 26029
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1050171852111816,
+      "learning_rate": 4.312186095691251e-05,
+      "loss": 1.4329,
+      "step": 26030
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.087703227996826,
+      "learning_rate": 4.3121596616441976e-05,
+      "loss": 1.3433,
+      "step": 26031
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8874174356460571,
+      "learning_rate": 4.3121332275971456e-05,
+      "loss": 1.2617,
+      "step": 26032
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6896135807037354,
+      "learning_rate": 4.312106793550092e-05,
+      "loss": 1.2469,
+      "step": 26033
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.4400343894958496,
+      "learning_rate": 4.3120803595030404e-05,
+      "loss": 1.426,
+      "step": 26034
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2822952270507812,
+      "learning_rate": 4.312053925455988e-05,
+      "loss": 2.0079,
+      "step": 26035
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.5047167539596558,
+      "learning_rate": 4.312027491408935e-05,
+      "loss": 1.5535,
+      "step": 26036
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6601184606552124,
+      "learning_rate": 4.3120010573618824e-05,
+      "loss": 1.1299,
+      "step": 26037
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0230822563171387,
+      "learning_rate": 4.31197462331483e-05,
+      "loss": 1.4822,
+      "step": 26038
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2986624240875244,
+      "learning_rate": 4.311948189267777e-05,
+      "loss": 1.2556,
+      "step": 26039
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8747082948684692,
+      "learning_rate": 4.3119217552207245e-05,
+      "loss": 1.2254,
+      "step": 26040
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2404661178588867,
+      "learning_rate": 4.311895321173672e-05,
+      "loss": 1.3146,
+      "step": 26041
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9194159507751465,
+      "learning_rate": 4.311868887126619e-05,
+      "loss": 1.2749,
+      "step": 26042
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8339554071426392,
+      "learning_rate": 4.3118424530795666e-05,
+      "loss": 1.2096,
+      "step": 26043
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.914000391960144,
+      "learning_rate": 4.311816019032514e-05,
+      "loss": 1.2693,
+      "step": 26044
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.916100263595581,
+      "learning_rate": 4.311789584985462e-05,
+      "loss": 1.4411,
+      "step": 26045
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9602166414260864,
+      "learning_rate": 4.3117631509384086e-05,
+      "loss": 1.2893,
+      "step": 26046
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6094944477081299,
+      "learning_rate": 4.311736716891357e-05,
+      "loss": 1.2813,
+      "step": 26047
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.987617015838623,
+      "learning_rate": 4.3117102828443034e-05,
+      "loss": 1.3465,
+      "step": 26048
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.3927993774414062,
+      "learning_rate": 4.3116838487972514e-05,
+      "loss": 2.2737,
+      "step": 26049
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9168124198913574,
+      "learning_rate": 4.311657414750199e-05,
+      "loss": 1.3536,
+      "step": 26050
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7587074041366577,
+      "learning_rate": 4.3116309807031454e-05,
+      "loss": 1.3824,
+      "step": 26051
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9148051738739014,
+      "learning_rate": 4.3116045466560935e-05,
+      "loss": 1.2375,
+      "step": 26052
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9008808135986328,
+      "learning_rate": 4.31157811260904e-05,
+      "loss": 1.2635,
+      "step": 26053
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.522979259490967,
+      "learning_rate": 4.311551678561988e-05,
+      "loss": 1.3521,
+      "step": 26054
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6875958442687988,
+      "learning_rate": 4.3115252445149355e-05,
+      "loss": 1.1454,
+      "step": 26055
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.072204351425171,
+      "learning_rate": 4.311498810467883e-05,
+      "loss": 1.3828,
+      "step": 26056
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9237632751464844,
+      "learning_rate": 4.31147237642083e-05,
+      "loss": 1.3236,
+      "step": 26057
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.187426567077637,
+      "learning_rate": 4.3114459423737776e-05,
+      "loss": 1.6415,
+      "step": 26058
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.5556323528289795,
+      "learning_rate": 4.311419508326725e-05,
+      "loss": 1.5148,
+      "step": 26059
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6837748289108276,
+      "learning_rate": 4.311393074279672e-05,
+      "loss": 1.1862,
+      "step": 26060
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.5655982494354248,
+      "learning_rate": 4.31136664023262e-05,
+      "loss": 1.1963,
+      "step": 26061
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8257783651351929,
+      "learning_rate": 4.311340206185567e-05,
+      "loss": 1.3243,
+      "step": 26062
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6689362525939941,
+      "learning_rate": 4.3113137721385144e-05,
+      "loss": 1.1917,
+      "step": 26063
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.810285210609436,
+      "learning_rate": 4.311287338091462e-05,
+      "loss": 1.2304,
+      "step": 26064
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.190514087677002,
+      "learning_rate": 4.31126090404441e-05,
+      "loss": 1.4726,
+      "step": 26065
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.17228364944458,
+      "learning_rate": 4.3112344699973565e-05,
+      "loss": 1.3119,
+      "step": 26066
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8502123355865479,
+      "learning_rate": 4.3112080359503045e-05,
+      "loss": 1.63,
+      "step": 26067
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.241687536239624,
+      "learning_rate": 4.311181601903251e-05,
+      "loss": 1.3181,
+      "step": 26068
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.5779669284820557,
+      "learning_rate": 4.311155167856199e-05,
+      "loss": 1.4328,
+      "step": 26069
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0335819721221924,
+      "learning_rate": 4.3111287338091466e-05,
+      "loss": 1.3769,
+      "step": 26070
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.640497922897339,
+      "learning_rate": 4.311102299762094e-05,
+      "loss": 1.3368,
+      "step": 26071
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8894977569580078,
+      "learning_rate": 4.311075865715041e-05,
+      "loss": 1.251,
+      "step": 26072
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8760112524032593,
+      "learning_rate": 4.3110494316679886e-05,
+      "loss": 1.2034,
+      "step": 26073
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.3903913497924805,
+      "learning_rate": 4.311022997620936e-05,
+      "loss": 1.5668,
+      "step": 26074
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.934760570526123,
+      "learning_rate": 4.3109965635738834e-05,
+      "loss": 1.398,
+      "step": 26075
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8876606225967407,
+      "learning_rate": 4.310970129526831e-05,
+      "loss": 1.371,
+      "step": 26076
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.800340175628662,
+      "learning_rate": 4.310943695479778e-05,
+      "loss": 1.7464,
+      "step": 26077
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.2927141189575195,
+      "learning_rate": 4.3109172614327254e-05,
+      "loss": 1.3469,
+      "step": 26078
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.465847969055176,
+      "learning_rate": 4.310890827385673e-05,
+      "loss": 1.1725,
+      "step": 26079
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.844729423522949,
+      "learning_rate": 4.310864393338621e-05,
+      "loss": 1.3164,
+      "step": 26080
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.871364593505859,
+      "learning_rate": 4.3108379592915675e-05,
+      "loss": 1.4162,
+      "step": 26081
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.095676898956299,
+      "learning_rate": 4.3108115252445155e-05,
+      "loss": 1.3333,
+      "step": 26082
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.040708303451538,
+      "learning_rate": 4.310785091197462e-05,
+      "loss": 1.2845,
+      "step": 26083
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.278984546661377,
+      "learning_rate": 4.31075865715041e-05,
+      "loss": 1.2079,
+      "step": 26084
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.328397274017334,
+      "learning_rate": 4.3107322231033576e-05,
+      "loss": 1.3283,
+      "step": 26085
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.6077115535736084,
+      "learning_rate": 4.310705789056305e-05,
+      "loss": 1.367,
+      "step": 26086
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0062568187713623,
+      "learning_rate": 4.310679355009252e-05,
+      "loss": 1.3024,
+      "step": 26087
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9726380109786987,
+      "learning_rate": 4.310652920962199e-05,
+      "loss": 1.3137,
+      "step": 26088
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.132728099822998,
+      "learning_rate": 4.310626486915147e-05,
+      "loss": 1.4627,
+      "step": 26089
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1847212314605713,
+      "learning_rate": 4.3106000528680944e-05,
+      "loss": 1.4616,
+      "step": 26090
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.403386354446411,
+      "learning_rate": 4.310573618821042e-05,
+      "loss": 1.4074,
+      "step": 26091
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2877938747406006,
+      "learning_rate": 4.310547184773989e-05,
+      "loss": 1.4466,
+      "step": 26092
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2906718254089355,
+      "learning_rate": 4.3105207507269365e-05,
+      "loss": 1.3311,
+      "step": 26093
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.005255937576294,
+      "learning_rate": 4.310494316679884e-05,
+      "loss": 1.2775,
+      "step": 26094
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0141501426696777,
+      "learning_rate": 4.310467882632832e-05,
+      "loss": 1.2706,
+      "step": 26095
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.3284800052642822,
+      "learning_rate": 4.3104414485857785e-05,
+      "loss": 1.3179,
+      "step": 26096
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0206236839294434,
+      "learning_rate": 4.310415014538726e-05,
+      "loss": 1.3073,
+      "step": 26097
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.482861042022705,
+      "learning_rate": 4.310388580491673e-05,
+      "loss": 1.225,
+      "step": 26098
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.264136791229248,
+      "learning_rate": 4.3103621464446206e-05,
+      "loss": 1.4044,
+      "step": 26099
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7112406492233276,
+      "learning_rate": 4.3103357123975686e-05,
+      "loss": 1.1865,
+      "step": 26100
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2206263542175293,
+      "learning_rate": 4.310309278350515e-05,
+      "loss": 1.439,
+      "step": 26101
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.698453426361084,
+      "learning_rate": 4.3102828443034633e-05,
+      "loss": 1.1685,
+      "step": 26102
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1929149627685547,
+      "learning_rate": 4.31025641025641e-05,
+      "loss": 1.4637,
+      "step": 26103
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9030920267105103,
+      "learning_rate": 4.310229976209358e-05,
+      "loss": 1.3281,
+      "step": 26104
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0059075355529785,
+      "learning_rate": 4.3102035421623054e-05,
+      "loss": 1.2326,
+      "step": 26105
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9250091314315796,
+      "learning_rate": 4.310177108115253e-05,
+      "loss": 1.3326,
+      "step": 26106
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.3780176639556885,
+      "learning_rate": 4.3101506740682e-05,
+      "loss": 1.3036,
+      "step": 26107
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.361020565032959,
+      "learning_rate": 4.3101242400211475e-05,
+      "loss": 1.4768,
+      "step": 26108
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8268327713012695,
+      "learning_rate": 4.310097805974095e-05,
+      "loss": 1.2693,
+      "step": 26109
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8097562789916992,
+      "learning_rate": 4.310071371927042e-05,
+      "loss": 1.3374,
+      "step": 26110
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9089043140411377,
+      "learning_rate": 4.3100449378799896e-05,
+      "loss": 1.2427,
+      "step": 26111
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.022862195968628,
+      "learning_rate": 4.310018503832937e-05,
+      "loss": 1.3062,
+      "step": 26112
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2126662731170654,
+      "learning_rate": 4.309992069785884e-05,
+      "loss": 1.098,
+      "step": 26113
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.820261836051941,
+      "learning_rate": 4.3099656357388316e-05,
+      "loss": 1.2503,
+      "step": 26114
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9071054458618164,
+      "learning_rate": 4.30993920169178e-05,
+      "loss": 1.3233,
+      "step": 26115
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.7650768756866455,
+      "learning_rate": 4.3099127676447263e-05,
+      "loss": 1.5063,
+      "step": 26116
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9554585218429565,
+      "learning_rate": 4.3098863335976744e-05,
+      "loss": 1.2597,
+      "step": 26117
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2839856147766113,
+      "learning_rate": 4.309859899550621e-05,
+      "loss": 1.2077,
+      "step": 26118
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7474660873413086,
+      "learning_rate": 4.309833465503569e-05,
+      "loss": 1.7106,
+      "step": 26119
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.017789125442505,
+      "learning_rate": 4.3098070314565165e-05,
+      "loss": 1.2897,
+      "step": 26120
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.734586715698242,
+      "learning_rate": 4.309780597409464e-05,
+      "loss": 1.4162,
+      "step": 26121
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0495972633361816,
+      "learning_rate": 4.309754163362411e-05,
+      "loss": 1.4014,
+      "step": 26122
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7857649326324463,
+      "learning_rate": 4.309727729315358e-05,
+      "loss": 1.2845,
+      "step": 26123
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9170236587524414,
+      "learning_rate": 4.309701295268306e-05,
+      "loss": 1.1873,
+      "step": 26124
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.060283899307251,
+      "learning_rate": 4.309674861221253e-05,
+      "loss": 1.2667,
+      "step": 26125
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.5931198596954346,
+      "learning_rate": 4.3096484271742006e-05,
+      "loss": 1.2949,
+      "step": 26126
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.8018436431884766,
+      "learning_rate": 4.309621993127148e-05,
+      "loss": 1.4706,
+      "step": 26127
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.026423692703247,
+      "learning_rate": 4.309595559080095e-05,
+      "loss": 1.3036,
+      "step": 26128
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9911824464797974,
+      "learning_rate": 4.309569125033043e-05,
+      "loss": 1.7742,
+      "step": 26129
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8330490589141846,
+      "learning_rate": 4.309542690985991e-05,
+      "loss": 1.2885,
+      "step": 26130
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.188476085662842,
+      "learning_rate": 4.3095162569389374e-05,
+      "loss": 1.3984,
+      "step": 26131
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1989333629608154,
+      "learning_rate": 4.309489822891885e-05,
+      "loss": 1.2843,
+      "step": 26132
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8264906406402588,
+      "learning_rate": 4.309463388844832e-05,
+      "loss": 1.2913,
+      "step": 26133
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.853994607925415,
+      "learning_rate": 4.3094369547977795e-05,
+      "loss": 1.2631,
+      "step": 26134
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8786424398422241,
+      "learning_rate": 4.3094105207507275e-05,
+      "loss": 1.2449,
+      "step": 26135
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1093263626098633,
+      "learning_rate": 4.309384086703674e-05,
+      "loss": 1.283,
+      "step": 26136
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7840198278427124,
+      "learning_rate": 4.309357652656622e-05,
+      "loss": 2.2776,
+      "step": 26137
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0070691108703613,
+      "learning_rate": 4.309331218609569e-05,
+      "loss": 1.3288,
+      "step": 26138
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6749963760375977,
+      "learning_rate": 4.309304784562517e-05,
+      "loss": 1.2721,
+      "step": 26139
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9271703958511353,
+      "learning_rate": 4.309278350515464e-05,
+      "loss": 1.3251,
+      "step": 26140
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.448735475540161,
+      "learning_rate": 4.3092519164684116e-05,
+      "loss": 1.3375,
+      "step": 26141
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.554663896560669,
+      "learning_rate": 4.309225482421359e-05,
+      "loss": 1.4311,
+      "step": 26142
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.353858709335327,
+      "learning_rate": 4.3091990483743063e-05,
+      "loss": 1.4952,
+      "step": 26143
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8365821838378906,
+      "learning_rate": 4.309172614327254e-05,
+      "loss": 1.2737,
+      "step": 26144
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.770292043685913,
+      "learning_rate": 4.309146180280201e-05,
+      "loss": 1.2707,
+      "step": 26145
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.7884175777435303,
+      "learning_rate": 4.3091197462331484e-05,
+      "loss": 1.36,
+      "step": 26146
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7012691497802734,
+      "learning_rate": 4.309093312186096e-05,
+      "loss": 1.2544,
+      "step": 26147
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7034261226654053,
+      "learning_rate": 4.309066878139043e-05,
+      "loss": 1.2626,
+      "step": 26148
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9893386363983154,
+      "learning_rate": 4.3090404440919905e-05,
+      "loss": 1.4103,
+      "step": 26149
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0927555561065674,
+      "learning_rate": 4.3090140100449385e-05,
+      "loss": 1.4386,
+      "step": 26150
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1051220893859863,
+      "learning_rate": 4.308987575997885e-05,
+      "loss": 1.2735,
+      "step": 26151
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.3983919620513916,
+      "learning_rate": 4.308961141950833e-05,
+      "loss": 1.3751,
+      "step": 26152
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.7051098346710205,
+      "learning_rate": 4.30893470790378e-05,
+      "loss": 1.4172,
+      "step": 26153
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8674503564834595,
+      "learning_rate": 4.308908273856728e-05,
+      "loss": 1.3173,
+      "step": 26154
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1811370849609375,
+      "learning_rate": 4.308881839809675e-05,
+      "loss": 1.3733,
+      "step": 26155
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.08886981010437,
+      "learning_rate": 4.308855405762623e-05,
+      "loss": 1.3178,
+      "step": 26156
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.5444447994232178,
+      "learning_rate": 4.30882897171557e-05,
+      "loss": 1.3342,
+      "step": 26157
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.4383418560028076,
+      "learning_rate": 4.3088025376685174e-05,
+      "loss": 1.2638,
+      "step": 26158
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9007902145385742,
+      "learning_rate": 4.308776103621465e-05,
+      "loss": 1.2919,
+      "step": 26159
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9432049989700317,
+      "learning_rate": 4.308749669574412e-05,
+      "loss": 1.1902,
+      "step": 26160
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9730124473571777,
+      "learning_rate": 4.3087232355273595e-05,
+      "loss": 1.2703,
+      "step": 26161
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7785643339157104,
+      "learning_rate": 4.308696801480307e-05,
+      "loss": 1.2252,
+      "step": 26162
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.846051812171936,
+      "learning_rate": 4.308670367433254e-05,
+      "loss": 1.3854,
+      "step": 26163
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7249584197998047,
+      "learning_rate": 4.3086439333862015e-05,
+      "loss": 1.2128,
+      "step": 26164
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.936490774154663,
+      "learning_rate": 4.3086174993391496e-05,
+      "loss": 1.3487,
+      "step": 26165
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.543368339538574,
+      "learning_rate": 4.308591065292096e-05,
+      "loss": 1.491,
+      "step": 26166
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1763527393341064,
+      "learning_rate": 4.308564631245044e-05,
+      "loss": 1.2901,
+      "step": 26167
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8570950031280518,
+      "learning_rate": 4.308538197197991e-05,
+      "loss": 1.3336,
+      "step": 26168
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.7598910331726074,
+      "learning_rate": 4.308511763150938e-05,
+      "loss": 1.4193,
+      "step": 26169
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8106963634490967,
+      "learning_rate": 4.3084853291038863e-05,
+      "loss": 1.0536,
+      "step": 26170
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0331594944000244,
+      "learning_rate": 4.308458895056833e-05,
+      "loss": 1.3839,
+      "step": 26171
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8715938329696655,
+      "learning_rate": 4.308432461009781e-05,
+      "loss": 1.3244,
+      "step": 26172
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8670001029968262,
+      "learning_rate": 4.308406026962728e-05,
+      "loss": 1.4401,
+      "step": 26173
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2923295497894287,
+      "learning_rate": 4.308379592915676e-05,
+      "loss": 1.3477,
+      "step": 26174
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9799467325210571,
+      "learning_rate": 4.308353158868623e-05,
+      "loss": 1.3387,
+      "step": 26175
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8333947658538818,
+      "learning_rate": 4.3083267248215705e-05,
+      "loss": 1.2581,
+      "step": 26176
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.091228485107422,
+      "learning_rate": 4.308300290774518e-05,
+      "loss": 1.3031,
+      "step": 26177
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2509584426879883,
+      "learning_rate": 4.308273856727465e-05,
+      "loss": 1.4211,
+      "step": 26178
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7819843292236328,
+      "learning_rate": 4.3082474226804126e-05,
+      "loss": 1.2294,
+      "step": 26179
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9661792516708374,
+      "learning_rate": 4.30822098863336e-05,
+      "loss": 1.3684,
+      "step": 26180
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.106717586517334,
+      "learning_rate": 4.308194554586307e-05,
+      "loss": 1.3283,
+      "step": 26181
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.8316383361816406,
+      "learning_rate": 4.3081681205392546e-05,
+      "loss": 1.2746,
+      "step": 26182
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.936221718788147,
+      "learning_rate": 4.308141686492202e-05,
+      "loss": 1.2609,
+      "step": 26183
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2793779373168945,
+      "learning_rate": 4.3081152524451493e-05,
+      "loss": 1.3543,
+      "step": 26184
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2272491455078125,
+      "learning_rate": 4.3080888183980974e-05,
+      "loss": 1.3354,
+      "step": 26185
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0837063789367676,
+      "learning_rate": 4.308062384351044e-05,
+      "loss": 1.3174,
+      "step": 26186
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9172037839889526,
+      "learning_rate": 4.308035950303992e-05,
+      "loss": 1.2186,
+      "step": 26187
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8847426176071167,
+      "learning_rate": 4.308009516256939e-05,
+      "loss": 1.2863,
+      "step": 26188
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6922478675842285,
+      "learning_rate": 4.307983082209887e-05,
+      "loss": 1.4251,
+      "step": 26189
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7273247241973877,
+      "learning_rate": 4.307956648162834e-05,
+      "loss": 1.2857,
+      "step": 26190
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7607135772705078,
+      "learning_rate": 4.3079302141157815e-05,
+      "loss": 1.2293,
+      "step": 26191
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.138058662414551,
+      "learning_rate": 4.307903780068729e-05,
+      "loss": 1.3926,
+      "step": 26192
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9478278160095215,
+      "learning_rate": 4.307877346021676e-05,
+      "loss": 1.271,
+      "step": 26193
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.209977388381958,
+      "learning_rate": 4.3078509119746236e-05,
+      "loss": 1.2841,
+      "step": 26194
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.5597587823867798,
+      "learning_rate": 4.307824477927571e-05,
+      "loss": 1.135,
+      "step": 26195
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.260946035385132,
+      "learning_rate": 4.307798043880518e-05,
+      "loss": 1.3565,
+      "step": 26196
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7746225595474243,
+      "learning_rate": 4.3077716098334657e-05,
+      "loss": 1.2134,
+      "step": 26197
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0324883460998535,
+      "learning_rate": 4.307745175786413e-05,
+      "loss": 1.2469,
+      "step": 26198
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.5901331901550293,
+      "learning_rate": 4.3077187417393604e-05,
+      "loss": 1.5066,
+      "step": 26199
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1153366565704346,
+      "learning_rate": 4.3076923076923084e-05,
+      "loss": 1.3519,
+      "step": 26200
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.426149368286133,
+      "learning_rate": 4.307665873645255e-05,
+      "loss": 1.4155,
+      "step": 26201
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.053311586380005,
+      "learning_rate": 4.307639439598203e-05,
+      "loss": 1.3529,
+      "step": 26202
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9179489612579346,
+      "learning_rate": 4.30761300555115e-05,
+      "loss": 1.3399,
+      "step": 26203
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.645890235900879,
+      "learning_rate": 4.307586571504098e-05,
+      "loss": 1.1687,
+      "step": 26204
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8739219903945923,
+      "learning_rate": 4.307560137457045e-05,
+      "loss": 1.6354,
+      "step": 26205
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.5374655723571777,
+      "learning_rate": 4.307533703409992e-05,
+      "loss": 1.2707,
+      "step": 26206
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9667850732803345,
+      "learning_rate": 4.30750726936294e-05,
+      "loss": 1.4175,
+      "step": 26207
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.5880870819091797,
+      "learning_rate": 4.3074808353158866e-05,
+      "loss": 1.3226,
+      "step": 26208
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.835975170135498,
+      "learning_rate": 4.3074544012688346e-05,
+      "loss": 1.4703,
+      "step": 26209
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0383379459381104,
+      "learning_rate": 4.307427967221782e-05,
+      "loss": 1.2274,
+      "step": 26210
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8976678848266602,
+      "learning_rate": 4.307401533174729e-05,
+      "loss": 1.3365,
+      "step": 26211
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9689836502075195,
+      "learning_rate": 4.307375099127677e-05,
+      "loss": 1.252,
+      "step": 26212
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0631937980651855,
+      "learning_rate": 4.307348665080624e-05,
+      "loss": 1.3099,
+      "step": 26213
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7404547929763794,
+      "learning_rate": 4.3073222310335714e-05,
+      "loss": 1.1493,
+      "step": 26214
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.4141054153442383,
+      "learning_rate": 4.307295796986519e-05,
+      "loss": 1.4636,
+      "step": 26215
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.83636474609375,
+      "learning_rate": 4.307269362939466e-05,
+      "loss": 1.3446,
+      "step": 26216
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2707674503326416,
+      "learning_rate": 4.3072429288924135e-05,
+      "loss": 1.3794,
+      "step": 26217
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0986886024475098,
+      "learning_rate": 4.307216494845361e-05,
+      "loss": 1.2597,
+      "step": 26218
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2194035053253174,
+      "learning_rate": 4.307190060798308e-05,
+      "loss": 1.736,
+      "step": 26219
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.823795795440674,
+      "learning_rate": 4.307163626751256e-05,
+      "loss": 1.4679,
+      "step": 26220
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8638077974319458,
+      "learning_rate": 4.307137192704203e-05,
+      "loss": 1.6596,
+      "step": 26221
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7940751314163208,
+      "learning_rate": 4.307110758657151e-05,
+      "loss": 1.0693,
+      "step": 26222
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9656070470809937,
+      "learning_rate": 4.3070843246100976e-05,
+      "loss": 1.2009,
+      "step": 26223
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1280734539031982,
+      "learning_rate": 4.3070578905630457e-05,
+      "loss": 1.2134,
+      "step": 26224
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8515574932098389,
+      "learning_rate": 4.307031456515993e-05,
+      "loss": 1.2493,
+      "step": 26225
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0798532962799072,
+      "learning_rate": 4.3070050224689404e-05,
+      "loss": 1.3199,
+      "step": 26226
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1033129692077637,
+      "learning_rate": 4.306978588421888e-05,
+      "loss": 1.3889,
+      "step": 26227
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1724510192871094,
+      "learning_rate": 4.306952154374835e-05,
+      "loss": 1.2072,
+      "step": 26228
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.122758150100708,
+      "learning_rate": 4.3069257203277824e-05,
+      "loss": 1.2666,
+      "step": 26229
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9682822227478027,
+      "learning_rate": 4.30689928628073e-05,
+      "loss": 1.2312,
+      "step": 26230
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8177673816680908,
+      "learning_rate": 4.306872852233677e-05,
+      "loss": 1.2216,
+      "step": 26231
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.876780390739441,
+      "learning_rate": 4.3068464181866245e-05,
+      "loss": 1.2963,
+      "step": 26232
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2669692039489746,
+      "learning_rate": 4.306819984139572e-05,
+      "loss": 1.2557,
+      "step": 26233
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.430353879928589,
+      "learning_rate": 4.306793550092519e-05,
+      "loss": 1.2201,
+      "step": 26234
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.651754379272461,
+      "learning_rate": 4.306767116045467e-05,
+      "loss": 1.3624,
+      "step": 26235
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.274430513381958,
+      "learning_rate": 4.306740681998414e-05,
+      "loss": 1.3636,
+      "step": 26236
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7825908660888672,
+      "learning_rate": 4.306714247951362e-05,
+      "loss": 1.0902,
+      "step": 26237
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2712507247924805,
+      "learning_rate": 4.3066878139043087e-05,
+      "loss": 1.29,
+      "step": 26238
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.7911460399627686,
+      "learning_rate": 4.306661379857257e-05,
+      "loss": 1.5552,
+      "step": 26239
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.924031376838684,
+      "learning_rate": 4.306634945810204e-05,
+      "loss": 1.232,
+      "step": 26240
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.5308117866516113,
+      "learning_rate": 4.306608511763151e-05,
+      "loss": 1.4766,
+      "step": 26241
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.343329668045044,
+      "learning_rate": 4.306582077716099e-05,
+      "loss": 1.4533,
+      "step": 26242
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.309877634048462,
+      "learning_rate": 4.3065556436690454e-05,
+      "loss": 1.4294,
+      "step": 26243
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0958425998687744,
+      "learning_rate": 4.3065292096219935e-05,
+      "loss": 1.3181,
+      "step": 26244
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1552388668060303,
+      "learning_rate": 4.306502775574941e-05,
+      "loss": 1.2977,
+      "step": 26245
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9323493242263794,
+      "learning_rate": 4.306476341527888e-05,
+      "loss": 1.1823,
+      "step": 26246
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0776820182800293,
+      "learning_rate": 4.3064499074808355e-05,
+      "loss": 1.2731,
+      "step": 26247
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1333110332489014,
+      "learning_rate": 4.306423473433783e-05,
+      "loss": 1.2514,
+      "step": 26248
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7926228046417236,
+      "learning_rate": 4.30639703938673e-05,
+      "loss": 1.2673,
+      "step": 26249
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1842703819274902,
+      "learning_rate": 4.3063706053396776e-05,
+      "loss": 1.3973,
+      "step": 26250
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.2712621688842773,
+      "learning_rate": 4.306344171292625e-05,
+      "loss": 1.5193,
+      "step": 26251
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8839982748031616,
+      "learning_rate": 4.306317737245572e-05,
+      "loss": 1.3161,
+      "step": 26252
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0748326778411865,
+      "learning_rate": 4.30629130319852e-05,
+      "loss": 1.2973,
+      "step": 26253
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.203310966491699,
+      "learning_rate": 4.306264869151467e-05,
+      "loss": 1.3798,
+      "step": 26254
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8760850429534912,
+      "learning_rate": 4.306238435104415e-05,
+      "loss": 1.3094,
+      "step": 26255
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8508069515228271,
+      "learning_rate": 4.306212001057362e-05,
+      "loss": 1.365,
+      "step": 26256
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.042802572250366,
+      "learning_rate": 4.30618556701031e-05,
+      "loss": 1.2688,
+      "step": 26257
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.29067325592041,
+      "learning_rate": 4.3061591329632565e-05,
+      "loss": 1.3767,
+      "step": 26258
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8101811408996582,
+      "learning_rate": 4.3061326989162045e-05,
+      "loss": 1.339,
+      "step": 26259
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1695075035095215,
+      "learning_rate": 4.306106264869152e-05,
+      "loss": 1.1786,
+      "step": 26260
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.192028760910034,
+      "learning_rate": 4.306079830822099e-05,
+      "loss": 1.3148,
+      "step": 26261
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.029139518737793,
+      "learning_rate": 4.3060533967750466e-05,
+      "loss": 1.3098,
+      "step": 26262
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.3131461143493652,
+      "learning_rate": 4.306026962727994e-05,
+      "loss": 1.4596,
+      "step": 26263
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.6955955028533936,
+      "learning_rate": 4.306000528680941e-05,
+      "loss": 1.239,
+      "step": 26264
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.4287109375,
+      "learning_rate": 4.3059740946338887e-05,
+      "loss": 1.2518,
+      "step": 26265
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0190348625183105,
+      "learning_rate": 4.305947660586836e-05,
+      "loss": 1.186,
+      "step": 26266
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8596653938293457,
+      "learning_rate": 4.3059212265397834e-05,
+      "loss": 1.1972,
+      "step": 26267
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8059289455413818,
+      "learning_rate": 4.305894792492731e-05,
+      "loss": 1.3819,
+      "step": 26268
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.7568676471710205,
+      "learning_rate": 4.305868358445678e-05,
+      "loss": 1.3606,
+      "step": 26269
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.197000503540039,
+      "learning_rate": 4.305841924398626e-05,
+      "loss": 1.3147,
+      "step": 26270
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.662680149078369,
+      "learning_rate": 4.305815490351573e-05,
+      "loss": 1.3212,
+      "step": 26271
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9839454889297485,
+      "learning_rate": 4.305789056304521e-05,
+      "loss": 1.259,
+      "step": 26272
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2915592193603516,
+      "learning_rate": 4.3057626222574675e-05,
+      "loss": 1.3227,
+      "step": 26273
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.699300527572632,
+      "learning_rate": 4.3057361882104155e-05,
+      "loss": 1.555,
+      "step": 26274
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.216714382171631,
+      "learning_rate": 4.305709754163363e-05,
+      "loss": 1.3595,
+      "step": 26275
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.3436660766601562,
+      "learning_rate": 4.30568332011631e-05,
+      "loss": 1.4175,
+      "step": 26276
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.965406060218811,
+      "learning_rate": 4.3056568860692576e-05,
+      "loss": 1.2554,
+      "step": 26277
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7588346004486084,
+      "learning_rate": 4.305630452022204e-05,
+      "loss": 1.1686,
+      "step": 26278
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7997714281082153,
+      "learning_rate": 4.305604017975152e-05,
+      "loss": 1.7089,
+      "step": 26279
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0833232402801514,
+      "learning_rate": 4.3055775839281e-05,
+      "loss": 1.3549,
+      "step": 26280
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9921138286590576,
+      "learning_rate": 4.305551149881047e-05,
+      "loss": 1.1787,
+      "step": 26281
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9100027084350586,
+      "learning_rate": 4.3055247158339944e-05,
+      "loss": 1.3729,
+      "step": 26282
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2038557529449463,
+      "learning_rate": 4.305498281786942e-05,
+      "loss": 1.3818,
+      "step": 26283
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.6634714603424072,
+      "learning_rate": 4.305471847739889e-05,
+      "loss": 1.4662,
+      "step": 26284
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.166822910308838,
+      "learning_rate": 4.305445413692837e-05,
+      "loss": 1.3711,
+      "step": 26285
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9334992170333862,
+      "learning_rate": 4.305418979645784e-05,
+      "loss": 1.2172,
+      "step": 26286
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1837475299835205,
+      "learning_rate": 4.305392545598731e-05,
+      "loss": 1.4625,
+      "step": 26287
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.0852749347686768,
+      "learning_rate": 4.3053661115516785e-05,
+      "loss": 1.3398,
+      "step": 26288
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.284991979598999,
+      "learning_rate": 4.305339677504626e-05,
+      "loss": 1.4286,
+      "step": 26289
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7956161499023438,
+      "learning_rate": 4.305313243457574e-05,
+      "loss": 1.1389,
+      "step": 26290
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.617662191390991,
+      "learning_rate": 4.3052868094105206e-05,
+      "loss": 1.2959,
+      "step": 26291
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.123640537261963,
+      "learning_rate": 4.3052603753634687e-05,
+      "loss": 1.2959,
+      "step": 26292
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.5422179698944092,
+      "learning_rate": 4.305233941316415e-05,
+      "loss": 1.214,
+      "step": 26293
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9612311124801636,
+      "learning_rate": 4.3052075072693634e-05,
+      "loss": 1.2164,
+      "step": 26294
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.3421614170074463,
+      "learning_rate": 4.305181073222311e-05,
+      "loss": 1.3703,
+      "step": 26295
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.894092321395874,
+      "learning_rate": 4.305154639175258e-05,
+      "loss": 1.2926,
+      "step": 26296
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1801559925079346,
+      "learning_rate": 4.3051282051282054e-05,
+      "loss": 1.3104,
+      "step": 26297
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.061631441116333,
+      "learning_rate": 4.305101771081153e-05,
+      "loss": 1.2761,
+      "step": 26298
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.631286859512329,
+      "learning_rate": 4.3050753370341e-05,
+      "loss": 1.5978,
+      "step": 26299
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9801952838897705,
+      "learning_rate": 4.3050489029870475e-05,
+      "loss": 1.2921,
+      "step": 26300
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2449469566345215,
+      "learning_rate": 4.305022468939995e-05,
+      "loss": 1.3477,
+      "step": 26301
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0188302993774414,
+      "learning_rate": 4.304996034892942e-05,
+      "loss": 1.2748,
+      "step": 26302
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.645086407661438,
+      "learning_rate": 4.3049696008458896e-05,
+      "loss": 2.0462,
+      "step": 26303
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.56901478767395,
+      "learning_rate": 4.304943166798837e-05,
+      "loss": 1.168,
+      "step": 26304
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.933371663093567,
+      "learning_rate": 4.304916732751785e-05,
+      "loss": 1.3075,
+      "step": 26305
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0711276531219482,
+      "learning_rate": 4.3048902987047316e-05,
+      "loss": 1.4144,
+      "step": 26306
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.6469616889953613,
+      "learning_rate": 4.30486386465768e-05,
+      "loss": 1.1209,
+      "step": 26307
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.139082670211792,
+      "learning_rate": 4.3048374306106264e-05,
+      "loss": 1.4079,
+      "step": 26308
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9638350009918213,
+      "learning_rate": 4.3048109965635744e-05,
+      "loss": 1.2424,
+      "step": 26309
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0890114307403564,
+      "learning_rate": 4.304784562516522e-05,
+      "loss": 1.3157,
+      "step": 26310
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.3911221027374268,
+      "learning_rate": 4.304758128469469e-05,
+      "loss": 1.4437,
+      "step": 26311
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.912619709968567,
+      "learning_rate": 4.3047316944224165e-05,
+      "loss": 1.2797,
+      "step": 26312
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0807957649230957,
+      "learning_rate": 4.304705260375363e-05,
+      "loss": 1.34,
+      "step": 26313
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.6393580436706543,
+      "learning_rate": 4.304678826328311e-05,
+      "loss": 1.5107,
+      "step": 26314
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7293615341186523,
+      "learning_rate": 4.3046523922812585e-05,
+      "loss": 1.2953,
+      "step": 26315
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9964076280593872,
+      "learning_rate": 4.304625958234206e-05,
+      "loss": 1.4654,
+      "step": 26316
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9289603233337402,
+      "learning_rate": 4.304599524187153e-05,
+      "loss": 1.0818,
+      "step": 26317
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.304501533508301,
+      "learning_rate": 4.3045730901401006e-05,
+      "loss": 1.2419,
+      "step": 26318
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.5744404792785645,
+      "learning_rate": 4.304546656093048e-05,
+      "loss": 1.4777,
+      "step": 26319
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8482061624526978,
+      "learning_rate": 4.304520222045996e-05,
+      "loss": 1.2821,
+      "step": 26320
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9041799306869507,
+      "learning_rate": 4.304493787998943e-05,
+      "loss": 1.2705,
+      "step": 26321
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9053977727890015,
+      "learning_rate": 4.304467353951891e-05,
+      "loss": 1.154,
+      "step": 26322
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9781310558319092,
+      "learning_rate": 4.3044409199048374e-05,
+      "loss": 1.2011,
+      "step": 26323
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.5567679405212402,
+      "learning_rate": 4.304414485857785e-05,
+      "loss": 1.1017,
+      "step": 26324
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9007172584533691,
+      "learning_rate": 4.304388051810733e-05,
+      "loss": 1.2964,
+      "step": 26325
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.689645528793335,
+      "learning_rate": 4.3043616177636795e-05,
+      "loss": 1.4166,
+      "step": 26326
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.365588665008545,
+      "learning_rate": 4.3043351837166275e-05,
+      "loss": 1.3261,
+      "step": 26327
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.5426502227783203,
+      "learning_rate": 4.304308749669574e-05,
+      "loss": 2.1513,
+      "step": 26328
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8566828966140747,
+      "learning_rate": 4.304282315622522e-05,
+      "loss": 1.3104,
+      "step": 26329
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.6167385578155518,
+      "learning_rate": 4.3042558815754696e-05,
+      "loss": 1.3218,
+      "step": 26330
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8697694540023804,
+      "learning_rate": 4.304229447528417e-05,
+      "loss": 1.2645,
+      "step": 26331
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1564900875091553,
+      "learning_rate": 4.304203013481364e-05,
+      "loss": 1.2412,
+      "step": 26332
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.4480252265930176,
+      "learning_rate": 4.3041765794343116e-05,
+      "loss": 1.406,
+      "step": 26333
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0286896228790283,
+      "learning_rate": 4.304150145387259e-05,
+      "loss": 1.3195,
+      "step": 26334
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8704094886779785,
+      "learning_rate": 4.3041237113402064e-05,
+      "loss": 1.2666,
+      "step": 26335
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.7759714126586914,
+      "learning_rate": 4.304097277293154e-05,
+      "loss": 1.2847,
+      "step": 26336
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8463239669799805,
+      "learning_rate": 4.304070843246101e-05,
+      "loss": 1.3908,
+      "step": 26337
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.5689616203308105,
+      "learning_rate": 4.3040444091990484e-05,
+      "loss": 1.3675,
+      "step": 26338
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.045126438140869,
+      "learning_rate": 4.304017975151996e-05,
+      "loss": 1.2925,
+      "step": 26339
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9551939964294434,
+      "learning_rate": 4.303991541104944e-05,
+      "loss": 1.367,
+      "step": 26340
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.337878704071045,
+      "learning_rate": 4.3039651070578905e-05,
+      "loss": 1.2769,
+      "step": 26341
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.3019418716430664,
+      "learning_rate": 4.3039386730108385e-05,
+      "loss": 1.4353,
+      "step": 26342
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8383821249008179,
+      "learning_rate": 4.303912238963785e-05,
+      "loss": 1.2116,
+      "step": 26343
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.893017053604126,
+      "learning_rate": 4.303885804916733e-05,
+      "loss": 1.3316,
+      "step": 26344
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2215778827667236,
+      "learning_rate": 4.3038593708696806e-05,
+      "loss": 1.2728,
+      "step": 26345
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.4559614658355713,
+      "learning_rate": 4.303832936822628e-05,
+      "loss": 1.4333,
+      "step": 26346
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.768537163734436,
+      "learning_rate": 4.303806502775575e-05,
+      "loss": 1.2862,
+      "step": 26347
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.75799560546875,
+      "learning_rate": 4.303780068728523e-05,
+      "loss": 1.201,
+      "step": 26348
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.160569906234741,
+      "learning_rate": 4.30375363468147e-05,
+      "loss": 1.3271,
+      "step": 26349
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.038537979125977,
+      "learning_rate": 4.3037272006344174e-05,
+      "loss": 2.5589,
+      "step": 26350
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.488229513168335,
+      "learning_rate": 4.303700766587365e-05,
+      "loss": 1.4557,
+      "step": 26351
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.348363161087036,
+      "learning_rate": 4.303674332540312e-05,
+      "loss": 1.3303,
+      "step": 26352
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2796437740325928,
+      "learning_rate": 4.3036478984932595e-05,
+      "loss": 1.3164,
+      "step": 26353
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.073040723800659,
+      "learning_rate": 4.303621464446207e-05,
+      "loss": 1.186,
+      "step": 26354
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0207667350769043,
+      "learning_rate": 4.303595030399155e-05,
+      "loss": 1.3485,
+      "step": 26355
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2336158752441406,
+      "learning_rate": 4.3035685963521015e-05,
+      "loss": 1.2837,
+      "step": 26356
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.330029249191284,
+      "learning_rate": 4.3035421623050496e-05,
+      "loss": 1.2842,
+      "step": 26357
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.368770122528076,
+      "learning_rate": 4.303515728257996e-05,
+      "loss": 1.471,
+      "step": 26358
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.6252409219741821,
+      "learning_rate": 4.3034892942109436e-05,
+      "loss": 1.2061,
+      "step": 26359
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.3170688152313232,
+      "learning_rate": 4.3034628601638916e-05,
+      "loss": 1.3806,
+      "step": 26360
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.012632131576538,
+      "learning_rate": 4.303436426116838e-05,
+      "loss": 1.3776,
+      "step": 26361
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9096581935882568,
+      "learning_rate": 4.3034099920697864e-05,
+      "loss": 1.352,
+      "step": 26362
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8835554122924805,
+      "learning_rate": 4.303383558022733e-05,
+      "loss": 1.2834,
+      "step": 26363
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.4011499881744385,
+      "learning_rate": 4.303357123975681e-05,
+      "loss": 2.3173,
+      "step": 26364
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.3351707458496094,
+      "learning_rate": 4.3033306899286284e-05,
+      "loss": 1.4763,
+      "step": 26365
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.3773903846740723,
+      "learning_rate": 4.303304255881576e-05,
+      "loss": 1.4562,
+      "step": 26366
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9616795778274536,
+      "learning_rate": 4.303277821834523e-05,
+      "loss": 1.3095,
+      "step": 26367
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.91744065284729,
+      "learning_rate": 4.3032513877874705e-05,
+      "loss": 1.2763,
+      "step": 26368
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8533223867416382,
+      "learning_rate": 4.303224953740418e-05,
+      "loss": 1.3758,
+      "step": 26369
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.1194400787353516,
+      "learning_rate": 4.303198519693365e-05,
+      "loss": 1.5212,
+      "step": 26370
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9833475351333618,
+      "learning_rate": 4.3031720856463126e-05,
+      "loss": 1.2757,
+      "step": 26371
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1879489421844482,
+      "learning_rate": 4.30314565159926e-05,
+      "loss": 1.3079,
+      "step": 26372
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9905201196670532,
+      "learning_rate": 4.303119217552207e-05,
+      "loss": 1.3681,
+      "step": 26373
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7172867059707642,
+      "learning_rate": 4.3030927835051546e-05,
+      "loss": 1.1933,
+      "step": 26374
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8295894861221313,
+      "learning_rate": 4.303066349458103e-05,
+      "loss": 1.2253,
+      "step": 26375
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.059577703475952,
+      "learning_rate": 4.3030399154110494e-05,
+      "loss": 1.2876,
+      "step": 26376
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.6435563564300537,
+      "learning_rate": 4.3030134813639974e-05,
+      "loss": 1.2664,
+      "step": 26377
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1000099182128906,
+      "learning_rate": 4.302987047316944e-05,
+      "loss": 1.2511,
+      "step": 26378
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.4429094791412354,
+      "learning_rate": 4.302960613269892e-05,
+      "loss": 1.4171,
+      "step": 26379
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.471713066101074,
+      "learning_rate": 4.3029341792228395e-05,
+      "loss": 1.371,
+      "step": 26380
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0944008827209473,
+      "learning_rate": 4.302907745175787e-05,
+      "loss": 1.2761,
+      "step": 26381
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.407116413116455,
+      "learning_rate": 4.302881311128734e-05,
+      "loss": 1.28,
+      "step": 26382
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1765944957733154,
+      "learning_rate": 4.3028548770816815e-05,
+      "loss": 1.3791,
+      "step": 26383
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.753570556640625,
+      "learning_rate": 4.302828443034629e-05,
+      "loss": 1.2405,
+      "step": 26384
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0014805793762207,
+      "learning_rate": 4.302802008987576e-05,
+      "loss": 1.2973,
+      "step": 26385
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.658446192741394,
+      "learning_rate": 4.3027755749405236e-05,
+      "loss": 1.2574,
+      "step": 26386
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.012721061706543,
+      "learning_rate": 4.302749140893471e-05,
+      "loss": 1.433,
+      "step": 26387
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1443543434143066,
+      "learning_rate": 4.302722706846418e-05,
+      "loss": 1.4064,
+      "step": 26388
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.225944995880127,
+      "learning_rate": 4.302696272799366e-05,
+      "loss": 1.2825,
+      "step": 26389
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.403580904006958,
+      "learning_rate": 4.302669838752314e-05,
+      "loss": 1.2158,
+      "step": 26390
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.5500857830047607,
+      "learning_rate": 4.3026434047052604e-05,
+      "loss": 1.3265,
+      "step": 26391
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6936771869659424,
+      "learning_rate": 4.3026169706582084e-05,
+      "loss": 1.229,
+      "step": 26392
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7479015588760376,
+      "learning_rate": 4.302590536611155e-05,
+      "loss": 1.094,
+      "step": 26393
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9610215425491333,
+      "learning_rate": 4.302564102564103e-05,
+      "loss": 1.1703,
+      "step": 26394
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.3925178050994873,
+      "learning_rate": 4.3025376685170505e-05,
+      "loss": 1.522,
+      "step": 26395
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.845028281211853,
+      "learning_rate": 4.302511234469997e-05,
+      "loss": 1.3111,
+      "step": 26396
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.816694974899292,
+      "learning_rate": 4.302484800422945e-05,
+      "loss": 1.4746,
+      "step": 26397
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.921244740486145,
+      "learning_rate": 4.302458366375892e-05,
+      "loss": 1.2906,
+      "step": 26398
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9086190462112427,
+      "learning_rate": 4.30243193232884e-05,
+      "loss": 1.0756,
+      "step": 26399
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.190478563308716,
+      "learning_rate": 4.302405498281787e-05,
+      "loss": 1.2083,
+      "step": 26400
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.5342246294021606,
+      "learning_rate": 4.3023790642347346e-05,
+      "loss": 1.2367,
+      "step": 26401
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.6118662357330322,
+      "learning_rate": 4.302352630187682e-05,
+      "loss": 1.5619,
+      "step": 26402
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.178172588348389,
+      "learning_rate": 4.3023261961406294e-05,
+      "loss": 1.3704,
+      "step": 26403
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9474939107894897,
+      "learning_rate": 4.302299762093577e-05,
+      "loss": 1.3347,
+      "step": 26404
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.055785655975342,
+      "learning_rate": 4.302273328046524e-05,
+      "loss": 1.3694,
+      "step": 26405
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.107245922088623,
+      "learning_rate": 4.3022468939994714e-05,
+      "loss": 1.4285,
+      "step": 26406
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.4056601524353027,
+      "learning_rate": 4.302220459952419e-05,
+      "loss": 1.3878,
+      "step": 26407
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.030519962310791,
+      "learning_rate": 4.302194025905366e-05,
+      "loss": 1.216,
+      "step": 26408
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3707852363586426,
+      "learning_rate": 4.3021675918583135e-05,
+      "loss": 1.3267,
+      "step": 26409
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.141530990600586,
+      "learning_rate": 4.3021411578112615e-05,
+      "loss": 1.2927,
+      "step": 26410
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8087142705917358,
+      "learning_rate": 4.302114723764208e-05,
+      "loss": 1.1436,
+      "step": 26411
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.4958975315093994,
+      "learning_rate": 4.302088289717156e-05,
+      "loss": 1.4117,
+      "step": 26412
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.947150707244873,
+      "learning_rate": 4.302061855670103e-05,
+      "loss": 1.3638,
+      "step": 26413
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.463189125061035,
+      "learning_rate": 4.302035421623051e-05,
+      "loss": 1.4498,
+      "step": 26414
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8944138288497925,
+      "learning_rate": 4.302008987575998e-05,
+      "loss": 1.1514,
+      "step": 26415
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.9472086429595947,
+      "learning_rate": 4.301982553528946e-05,
+      "loss": 1.7718,
+      "step": 26416
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3620853424072266,
+      "learning_rate": 4.301956119481893e-05,
+      "loss": 1.3008,
+      "step": 26417
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.111234426498413,
+      "learning_rate": 4.3019296854348404e-05,
+      "loss": 1.2554,
+      "step": 26418
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7112560272216797,
+      "learning_rate": 4.301903251387788e-05,
+      "loss": 1.2911,
+      "step": 26419
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.4382688999176025,
+      "learning_rate": 4.301876817340735e-05,
+      "loss": 1.7018,
+      "step": 26420
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.7632925510406494,
+      "learning_rate": 4.3018503832936825e-05,
+      "loss": 1.3465,
+      "step": 26421
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.179863452911377,
+      "learning_rate": 4.30182394924663e-05,
+      "loss": 1.4012,
+      "step": 26422
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9037787914276123,
+      "learning_rate": 4.301797515199577e-05,
+      "loss": 1.3634,
+      "step": 26423
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.7385449409484863,
+      "learning_rate": 4.3017710811525245e-05,
+      "loss": 1.396,
+      "step": 26424
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.952209234237671,
+      "learning_rate": 4.3017446471054726e-05,
+      "loss": 1.3263,
+      "step": 26425
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6098670959472656,
+      "learning_rate": 4.301718213058419e-05,
+      "loss": 1.2218,
+      "step": 26426
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0964653491973877,
+      "learning_rate": 4.301691779011367e-05,
+      "loss": 1.385,
+      "step": 26427
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.885838270187378,
+      "learning_rate": 4.301665344964314e-05,
+      "loss": 1.2542,
+      "step": 26428
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7508295774459839,
+      "learning_rate": 4.301638910917262e-05,
+      "loss": 1.2745,
+      "step": 26429
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3801498413085938,
+      "learning_rate": 4.3016124768702093e-05,
+      "loss": 1.3419,
+      "step": 26430
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9338949918746948,
+      "learning_rate": 4.301586042823156e-05,
+      "loss": 1.1878,
+      "step": 26431
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.002922534942627,
+      "learning_rate": 4.301559608776104e-05,
+      "loss": 1.2244,
+      "step": 26432
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8773006200790405,
+      "learning_rate": 4.301533174729051e-05,
+      "loss": 1.3363,
+      "step": 26433
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.948291301727295,
+      "learning_rate": 4.301506740681999e-05,
+      "loss": 1.3119,
+      "step": 26434
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3904666900634766,
+      "learning_rate": 4.301480306634946e-05,
+      "loss": 1.4124,
+      "step": 26435
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.2470176219940186,
+      "learning_rate": 4.3014538725878935e-05,
+      "loss": 1.2362,
+      "step": 26436
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.175332546234131,
+      "learning_rate": 4.301427438540841e-05,
+      "loss": 1.3656,
+      "step": 26437
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8108361959457397,
+      "learning_rate": 4.301401004493788e-05,
+      "loss": 1.3512,
+      "step": 26438
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9983035326004028,
+      "learning_rate": 4.3013745704467356e-05,
+      "loss": 1.3566,
+      "step": 26439
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6910178661346436,
+      "learning_rate": 4.301348136399683e-05,
+      "loss": 1.1963,
+      "step": 26440
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.144346237182617,
+      "learning_rate": 4.30132170235263e-05,
+      "loss": 1.3755,
+      "step": 26441
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.086017608642578,
+      "learning_rate": 4.3012952683055776e-05,
+      "loss": 1.2185,
+      "step": 26442
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.694429874420166,
+      "learning_rate": 4.301268834258525e-05,
+      "loss": 1.1868,
+      "step": 26443
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.201641082763672,
+      "learning_rate": 4.3012424002114723e-05,
+      "loss": 1.3723,
+      "step": 26444
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.09362530708313,
+      "learning_rate": 4.3012159661644204e-05,
+      "loss": 1.3379,
+      "step": 26445
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6110711097717285,
+      "learning_rate": 4.301189532117367e-05,
+      "loss": 1.1823,
+      "step": 26446
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.167849540710449,
+      "learning_rate": 4.301163098070315e-05,
+      "loss": 1.3234,
+      "step": 26447
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9578914642333984,
+      "learning_rate": 4.301136664023262e-05,
+      "loss": 1.3378,
+      "step": 26448
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.902877688407898,
+      "learning_rate": 4.30111022997621e-05,
+      "loss": 1.2093,
+      "step": 26449
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7444409132003784,
+      "learning_rate": 4.301083795929157e-05,
+      "loss": 1.5323,
+      "step": 26450
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.4823811054229736,
+      "learning_rate": 4.3010573618821045e-05,
+      "loss": 1.4791,
+      "step": 26451
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.116572618484497,
+      "learning_rate": 4.301030927835052e-05,
+      "loss": 1.4579,
+      "step": 26452
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.507296323776245,
+      "learning_rate": 4.301004493787999e-05,
+      "loss": 1.3818,
+      "step": 26453
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.416811943054199,
+      "learning_rate": 4.3009780597409466e-05,
+      "loss": 1.3285,
+      "step": 26454
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0323269367218018,
+      "learning_rate": 4.300951625693894e-05,
+      "loss": 1.3657,
+      "step": 26455
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0978739261627197,
+      "learning_rate": 4.300925191646841e-05,
+      "loss": 1.3846,
+      "step": 26456
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.841925024986267,
+      "learning_rate": 4.300898757599789e-05,
+      "loss": 1.2446,
+      "step": 26457
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8307267427444458,
+      "learning_rate": 4.300872323552736e-05,
+      "loss": 1.2786,
+      "step": 26458
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9208064079284668,
+      "learning_rate": 4.3008458895056834e-05,
+      "loss": 1.2638,
+      "step": 26459
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.903293251991272,
+      "learning_rate": 4.3008194554586314e-05,
+      "loss": 1.282,
+      "step": 26460
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0395612716674805,
+      "learning_rate": 4.300793021411578e-05,
+      "loss": 1.2721,
+      "step": 26461
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.2255632877349854,
+      "learning_rate": 4.300766587364526e-05,
+      "loss": 2.1743,
+      "step": 26462
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.0359697341918945,
+      "learning_rate": 4.300740153317473e-05,
+      "loss": 1.4252,
+      "step": 26463
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.1473522186279297,
+      "learning_rate": 4.300713719270421e-05,
+      "loss": 1.4256,
+      "step": 26464
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.121321201324463,
+      "learning_rate": 4.300687285223368e-05,
+      "loss": 1.0956,
+      "step": 26465
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8272939920425415,
+      "learning_rate": 4.3006608511763156e-05,
+      "loss": 1.2214,
+      "step": 26466
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.7887322902679443,
+      "learning_rate": 4.300634417129263e-05,
+      "loss": 1.4664,
+      "step": 26467
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.009436845779419,
+      "learning_rate": 4.3006079830822096e-05,
+      "loss": 1.239,
+      "step": 26468
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.561699151992798,
+      "learning_rate": 4.3005815490351576e-05,
+      "loss": 1.3578,
+      "step": 26469
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.092508316040039,
+      "learning_rate": 4.300555114988105e-05,
+      "loss": 1.2766,
+      "step": 26470
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9652154445648193,
+      "learning_rate": 4.3005286809410523e-05,
+      "loss": 1.4076,
+      "step": 26471
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.192678689956665,
+      "learning_rate": 4.300502246894e-05,
+      "loss": 1.3694,
+      "step": 26472
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0680654048919678,
+      "learning_rate": 4.300475812846947e-05,
+      "loss": 1.2767,
+      "step": 26473
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7351009845733643,
+      "learning_rate": 4.3004493787998944e-05,
+      "loss": 1.2924,
+      "step": 26474
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3712542057037354,
+      "learning_rate": 4.3004229447528425e-05,
+      "loss": 1.3974,
+      "step": 26475
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.1929733753204346,
+      "learning_rate": 4.300396510705789e-05,
+      "loss": 1.2672,
+      "step": 26476
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9805885553359985,
+      "learning_rate": 4.3003700766587365e-05,
+      "loss": 1.1048,
+      "step": 26477
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9432921409606934,
+      "learning_rate": 4.300343642611684e-05,
+      "loss": 1.3602,
+      "step": 26478
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.725373387336731,
+      "learning_rate": 4.300317208564631e-05,
+      "loss": 1.1303,
+      "step": 26479
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9684090614318848,
+      "learning_rate": 4.300290774517579e-05,
+      "loss": 1.2064,
+      "step": 26480
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.188701868057251,
+      "learning_rate": 4.300264340470526e-05,
+      "loss": 1.2306,
+      "step": 26481
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.212735652923584,
+      "learning_rate": 4.300237906423474e-05,
+      "loss": 1.2856,
+      "step": 26482
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.352973699569702,
+      "learning_rate": 4.3002114723764206e-05,
+      "loss": 1.4057,
+      "step": 26483
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.950759768486023,
+      "learning_rate": 4.300185038329369e-05,
+      "loss": 1.3536,
+      "step": 26484
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9832394123077393,
+      "learning_rate": 4.300158604282316e-05,
+      "loss": 1.1785,
+      "step": 26485
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7335619926452637,
+      "learning_rate": 4.3001321702352634e-05,
+      "loss": 1.1771,
+      "step": 26486
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0537304878234863,
+      "learning_rate": 4.300105736188211e-05,
+      "loss": 1.3316,
+      "step": 26487
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.1329569816589355,
+      "learning_rate": 4.300079302141158e-05,
+      "loss": 1.3638,
+      "step": 26488
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9617655277252197,
+      "learning_rate": 4.3000528680941055e-05,
+      "loss": 1.3064,
+      "step": 26489
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6395690441131592,
+      "learning_rate": 4.300026434047053e-05,
+      "loss": 1.4976,
+      "step": 26490
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7539414167404175,
+      "learning_rate": 4.3e-05,
+      "loss": 1.2431,
+      "step": 26491
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8444390296936035,
+      "learning_rate": 4.2999735659529475e-05,
+      "loss": 1.2629,
+      "step": 26492
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.230909585952759,
+      "learning_rate": 4.299947131905895e-05,
+      "loss": 1.6223,
+      "step": 26493
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.022996425628662,
+      "learning_rate": 4.299920697858842e-05,
+      "loss": 1.3807,
+      "step": 26494
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9958045482635498,
+      "learning_rate": 4.29989426381179e-05,
+      "loss": 1.3005,
+      "step": 26495
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.1357052326202393,
+      "learning_rate": 4.299867829764737e-05,
+      "loss": 1.3287,
+      "step": 26496
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.231328248977661,
+      "learning_rate": 4.299841395717685e-05,
+      "loss": 1.1572,
+      "step": 26497
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9951552152633667,
+      "learning_rate": 4.299814961670632e-05,
+      "loss": 1.2126,
+      "step": 26498
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.7355282306671143,
+      "learning_rate": 4.29978852762358e-05,
+      "loss": 1.3724,
+      "step": 26499
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.184455633163452,
+      "learning_rate": 4.299762093576527e-05,
+      "loss": 1.2155,
+      "step": 26500
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.611424684524536,
+      "learning_rate": 4.2997356595294744e-05,
+      "loss": 1.2954,
+      "step": 26501
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.011298179626465,
+      "learning_rate": 4.299709225482422e-05,
+      "loss": 1.8041,
+      "step": 26502
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9347745180130005,
+      "learning_rate": 4.2996827914353685e-05,
+      "loss": 1.2948,
+      "step": 26503
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.223538398742676,
+      "learning_rate": 4.2996563573883165e-05,
+      "loss": 1.3115,
+      "step": 26504
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.1335372924804688,
+      "learning_rate": 4.299629923341264e-05,
+      "loss": 1.4204,
+      "step": 26505
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3944640159606934,
+      "learning_rate": 4.299603489294211e-05,
+      "loss": 1.3346,
+      "step": 26506
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6915748119354248,
+      "learning_rate": 4.2995770552471586e-05,
+      "loss": 1.2549,
+      "step": 26507
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.2377378940582275,
+      "learning_rate": 4.299550621200106e-05,
+      "loss": 1.3212,
+      "step": 26508
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9503225088119507,
+      "learning_rate": 4.299524187153053e-05,
+      "loss": 1.342,
+      "step": 26509
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.223931312561035,
+      "learning_rate": 4.299497753106001e-05,
+      "loss": 1.3289,
+      "step": 26510
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3209474086761475,
+      "learning_rate": 4.299471319058948e-05,
+      "loss": 1.4038,
+      "step": 26511
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9556231498718262,
+      "learning_rate": 4.299444885011896e-05,
+      "loss": 1.2283,
+      "step": 26512
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9648116827011108,
+      "learning_rate": 4.299418450964843e-05,
+      "loss": 1.2541,
+      "step": 26513
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7845019102096558,
+      "learning_rate": 4.29939201691779e-05,
+      "loss": 1.1729,
+      "step": 26514
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.7707509994506836,
+      "learning_rate": 4.299365582870738e-05,
+      "loss": 1.3722,
+      "step": 26515
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.4730067253112793,
+      "learning_rate": 4.299339148823685e-05,
+      "loss": 2.4617,
+      "step": 26516
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9820706844329834,
+      "learning_rate": 4.299312714776633e-05,
+      "loss": 1.243,
+      "step": 26517
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7318718433380127,
+      "learning_rate": 4.2992862807295795e-05,
+      "loss": 1.2304,
+      "step": 26518
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.626976728439331,
+      "learning_rate": 4.2992598466825275e-05,
+      "loss": 1.131,
+      "step": 26519
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0153303146362305,
+      "learning_rate": 4.299233412635475e-05,
+      "loss": 1.1869,
+      "step": 26520
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.309297561645508,
+      "learning_rate": 4.299206978588422e-05,
+      "loss": 1.2927,
+      "step": 26521
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8456542491912842,
+      "learning_rate": 4.2991805445413696e-05,
+      "loss": 1.2785,
+      "step": 26522
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0997118949890137,
+      "learning_rate": 4.299154110494317e-05,
+      "loss": 1.3092,
+      "step": 26523
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.012134075164795,
+      "learning_rate": 4.299127676447264e-05,
+      "loss": 1.4184,
+      "step": 26524
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.6811530590057373,
+      "learning_rate": 4.299101242400212e-05,
+      "loss": 1.4522,
+      "step": 26525
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.1016793251037598,
+      "learning_rate": 4.299074808353159e-05,
+      "loss": 1.2633,
+      "step": 26526
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8972104787826538,
+      "learning_rate": 4.2990483743061064e-05,
+      "loss": 1.2537,
+      "step": 26527
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.1129467487335205,
+      "learning_rate": 4.299021940259054e-05,
+      "loss": 1.3962,
+      "step": 26528
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.4548304080963135,
+      "learning_rate": 4.298995506212001e-05,
+      "loss": 1.4396,
+      "step": 26529
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6702584028244019,
+      "learning_rate": 4.298969072164949e-05,
+      "loss": 1.2406,
+      "step": 26530
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9992563724517822,
+      "learning_rate": 4.298942638117896e-05,
+      "loss": 1.2698,
+      "step": 26531
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.9443633556365967,
+      "learning_rate": 4.298916204070844e-05,
+      "loss": 1.4192,
+      "step": 26532
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.1833879947662354,
+      "learning_rate": 4.2988897700237905e-05,
+      "loss": 1.4364,
+      "step": 26533
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.1717758178710938,
+      "learning_rate": 4.2988633359767386e-05,
+      "loss": 1.6246,
+      "step": 26534
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.831313967704773,
+      "learning_rate": 4.298836901929686e-05,
+      "loss": 1.2456,
+      "step": 26535
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.4680955410003662,
+      "learning_rate": 4.298810467882633e-05,
+      "loss": 1.2001,
+      "step": 26536
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.355623722076416,
+      "learning_rate": 4.2987840338355806e-05,
+      "loss": 1.2168,
+      "step": 26537
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.968000888824463,
+      "learning_rate": 4.298757599788528e-05,
+      "loss": 1.1948,
+      "step": 26538
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.230081796646118,
+      "learning_rate": 4.298731165741475e-05,
+      "loss": 1.598,
+      "step": 26539
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0957155227661133,
+      "learning_rate": 4.298704731694423e-05,
+      "loss": 1.2989,
+      "step": 26540
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8140294551849365,
+      "learning_rate": 4.29867829764737e-05,
+      "loss": 1.8363,
+      "step": 26541
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0644655227661133,
+      "learning_rate": 4.2986518636003174e-05,
+      "loss": 1.322,
+      "step": 26542
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.294900417327881,
+      "learning_rate": 4.298625429553265e-05,
+      "loss": 1.3673,
+      "step": 26543
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8212181329727173,
+      "learning_rate": 4.298598995506212e-05,
+      "loss": 1.3008,
+      "step": 26544
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.297886371612549,
+      "learning_rate": 4.29857256145916e-05,
+      "loss": 1.3231,
+      "step": 26545
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8890917301177979,
+      "learning_rate": 4.298546127412107e-05,
+      "loss": 1.2876,
+      "step": 26546
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.114773750305176,
+      "learning_rate": 4.298519693365055e-05,
+      "loss": 1.3915,
+      "step": 26547
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.359494924545288,
+      "learning_rate": 4.2984932593180016e-05,
+      "loss": 1.3607,
+      "step": 26548
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.4338316917419434,
+      "learning_rate": 4.298466825270949e-05,
+      "loss": 1.3185,
+      "step": 26549
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9346226453781128,
+      "learning_rate": 4.298440391223897e-05,
+      "loss": 1.2194,
+      "step": 26550
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9523216485977173,
+      "learning_rate": 4.2984139571768436e-05,
+      "loss": 1.1934,
+      "step": 26551
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6875572204589844,
+      "learning_rate": 4.2983875231297917e-05,
+      "loss": 1.2007,
+      "step": 26552
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3430349826812744,
+      "learning_rate": 4.298361089082738e-05,
+      "loss": 1.4993,
+      "step": 26553
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9968502521514893,
+      "learning_rate": 4.2983346550356864e-05,
+      "loss": 1.3777,
+      "step": 26554
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.366206407546997,
+      "learning_rate": 4.298308220988634e-05,
+      "loss": 1.3262,
+      "step": 26555
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8910490274429321,
+      "learning_rate": 4.298281786941581e-05,
+      "loss": 1.3601,
+      "step": 26556
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7347948551177979,
+      "learning_rate": 4.2982553528945284e-05,
+      "loss": 1.3157,
+      "step": 26557
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.7079384326934814,
+      "learning_rate": 4.298228918847476e-05,
+      "loss": 1.1883,
+      "step": 26558
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.2439801692962646,
+      "learning_rate": 4.298202484800423e-05,
+      "loss": 1.315,
+      "step": 26559
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7449440956115723,
+      "learning_rate": 4.2981760507533705e-05,
+      "loss": 1.293,
+      "step": 26560
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9343574047088623,
+      "learning_rate": 4.298149616706318e-05,
+      "loss": 1.1727,
+      "step": 26561
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.0785744190216064,
+      "learning_rate": 4.298123182659265e-05,
+      "loss": 1.3507,
+      "step": 26562
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.153533697128296,
+      "learning_rate": 4.2980967486122126e-05,
+      "loss": 1.3595,
+      "step": 26563
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3429510593414307,
+      "learning_rate": 4.29807031456516e-05,
+      "loss": 1.3888,
+      "step": 26564
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7173700332641602,
+      "learning_rate": 4.298043880518108e-05,
+      "loss": 1.4838,
+      "step": 26565
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3844544887542725,
+      "learning_rate": 4.2980174464710547e-05,
+      "loss": 1.3683,
+      "step": 26566
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.5064784288406372,
+      "learning_rate": 4.297991012424003e-05,
+      "loss": 3.093,
+      "step": 26567
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.5584065914154053,
+      "learning_rate": 4.2979645783769494e-05,
+      "loss": 1.4437,
+      "step": 26568
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8687132596969604,
+      "learning_rate": 4.2979381443298974e-05,
+      "loss": 2.208,
+      "step": 26569
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.6432547569274902,
+      "learning_rate": 4.297911710282845e-05,
+      "loss": 1.3205,
+      "step": 26570
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.9459006786346436,
+      "learning_rate": 4.297885276235792e-05,
+      "loss": 1.3562,
+      "step": 26571
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.760516881942749,
+      "learning_rate": 4.2978588421887395e-05,
+      "loss": 1.4116,
+      "step": 26572
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8095730543136597,
+      "learning_rate": 4.297832408141687e-05,
+      "loss": 1.1786,
+      "step": 26573
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6281532049179077,
+      "learning_rate": 4.297805974094634e-05,
+      "loss": 1.2678,
+      "step": 26574
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.4543516635894775,
+      "learning_rate": 4.2977795400475815e-05,
+      "loss": 1.4477,
+      "step": 26575
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.369220018386841,
+      "learning_rate": 4.297753106000529e-05,
+      "loss": 1.3303,
+      "step": 26576
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.8230024576187134,
+      "learning_rate": 4.297726671953476e-05,
+      "loss": 1.2722,
+      "step": 26577
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1294820308685303,
+      "learning_rate": 4.2977002379064236e-05,
+      "loss": 1.3007,
+      "step": 26578
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.083111524581909,
+      "learning_rate": 4.297673803859371e-05,
+      "loss": 1.2574,
+      "step": 26579
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9727673530578613,
+      "learning_rate": 4.297647369812319e-05,
+      "loss": 1.3534,
+      "step": 26580
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.933542013168335,
+      "learning_rate": 4.297620935765266e-05,
+      "loss": 1.3305,
+      "step": 26581
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.346853256225586,
+      "learning_rate": 4.297594501718214e-05,
+      "loss": 1.3905,
+      "step": 26582
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1422760486602783,
+      "learning_rate": 4.2975680676711604e-05,
+      "loss": 1.2495,
+      "step": 26583
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6520752906799316,
+      "learning_rate": 4.2975416336241084e-05,
+      "loss": 1.2684,
+      "step": 26584
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0073599815368652,
+      "learning_rate": 4.297515199577056e-05,
+      "loss": 1.4711,
+      "step": 26585
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.987257480621338,
+      "learning_rate": 4.2974887655300025e-05,
+      "loss": 1.2602,
+      "step": 26586
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6599438190460205,
+      "learning_rate": 4.2974623314829505e-05,
+      "loss": 1.2778,
+      "step": 26587
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8703489303588867,
+      "learning_rate": 4.297435897435897e-05,
+      "loss": 1.3226,
+      "step": 26588
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.785959243774414,
+      "learning_rate": 4.297409463388845e-05,
+      "loss": 1.228,
+      "step": 26589
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0749189853668213,
+      "learning_rate": 4.2973830293417926e-05,
+      "loss": 1.3429,
+      "step": 26590
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1124024391174316,
+      "learning_rate": 4.29735659529474e-05,
+      "loss": 1.2602,
+      "step": 26591
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.314234495162964,
+      "learning_rate": 4.297330161247687e-05,
+      "loss": 1.5435,
+      "step": 26592
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7535053491592407,
+      "learning_rate": 4.2973037272006347e-05,
+      "loss": 1.1922,
+      "step": 26593
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.958829402923584,
+      "learning_rate": 4.297277293153582e-05,
+      "loss": 1.5581,
+      "step": 26594
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.768580675125122,
+      "learning_rate": 4.2972508591065294e-05,
+      "loss": 1.1695,
+      "step": 26595
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.901012659072876,
+      "learning_rate": 4.297224425059477e-05,
+      "loss": 1.3293,
+      "step": 26596
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.003213882446289,
+      "learning_rate": 4.297197991012424e-05,
+      "loss": 1.209,
+      "step": 26597
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.3721561431884766,
+      "learning_rate": 4.2971715569653714e-05,
+      "loss": 1.5381,
+      "step": 26598
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.211994171142578,
+      "learning_rate": 4.297145122918319e-05,
+      "loss": 1.4723,
+      "step": 26599
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1108462810516357,
+      "learning_rate": 4.297118688871267e-05,
+      "loss": 1.3442,
+      "step": 26600
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6715166568756104,
+      "learning_rate": 4.2970922548242135e-05,
+      "loss": 1.2037,
+      "step": 26601
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.146188735961914,
+      "learning_rate": 4.2970658207771615e-05,
+      "loss": 1.3269,
+      "step": 26602
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.01495623588562,
+      "learning_rate": 4.297039386730108e-05,
+      "loss": 1.3154,
+      "step": 26603
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.726597547531128,
+      "learning_rate": 4.297012952683056e-05,
+      "loss": 1.2823,
+      "step": 26604
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0223395824432373,
+      "learning_rate": 4.2969865186360036e-05,
+      "loss": 1.2698,
+      "step": 26605
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0140790939331055,
+      "learning_rate": 4.296960084588951e-05,
+      "loss": 1.3187,
+      "step": 26606
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.26279616355896,
+      "learning_rate": 4.296933650541898e-05,
+      "loss": 1.3622,
+      "step": 26607
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2631547451019287,
+      "learning_rate": 4.296907216494846e-05,
+      "loss": 1.401,
+      "step": 26608
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.3348388671875,
+      "learning_rate": 4.296880782447793e-05,
+      "loss": 1.3735,
+      "step": 26609
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8308649063110352,
+      "learning_rate": 4.2968543484007404e-05,
+      "loss": 1.2436,
+      "step": 26610
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8172599077224731,
+      "learning_rate": 4.296827914353688e-05,
+      "loss": 1.3055,
+      "step": 26611
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.101477861404419,
+      "learning_rate": 4.296801480306635e-05,
+      "loss": 1.621,
+      "step": 26612
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8581289052963257,
+      "learning_rate": 4.2967750462595825e-05,
+      "loss": 1.2243,
+      "step": 26613
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0574183464050293,
+      "learning_rate": 4.29674861221253e-05,
+      "loss": 1.2889,
+      "step": 26614
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.86844801902771,
+      "learning_rate": 4.296722178165477e-05,
+      "loss": 1.2227,
+      "step": 26615
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.122964859008789,
+      "learning_rate": 4.2966957441184245e-05,
+      "loss": 1.3673,
+      "step": 26616
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.364560604095459,
+      "learning_rate": 4.2966693100713726e-05,
+      "loss": 1.2243,
+      "step": 26617
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2682013511657715,
+      "learning_rate": 4.296642876024319e-05,
+      "loss": 1.3447,
+      "step": 26618
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.217125415802002,
+      "learning_rate": 4.296616441977267e-05,
+      "loss": 1.4158,
+      "step": 26619
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.473158836364746,
+      "learning_rate": 4.296590007930214e-05,
+      "loss": 1.3048,
+      "step": 26620
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8560097217559814,
+      "learning_rate": 4.296563573883161e-05,
+      "loss": 1.2868,
+      "step": 26621
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.4663894176483154,
+      "learning_rate": 4.2965371398361094e-05,
+      "loss": 1.3236,
+      "step": 26622
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8348315954208374,
+      "learning_rate": 4.296510705789056e-05,
+      "loss": 1.3546,
+      "step": 26623
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.042140245437622,
+      "learning_rate": 4.296484271742004e-05,
+      "loss": 1.2804,
+      "step": 26624
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9901494979858398,
+      "learning_rate": 4.296457837694951e-05,
+      "loss": 1.2755,
+      "step": 26625
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.744375228881836,
+      "learning_rate": 4.296431403647899e-05,
+      "loss": 1.1867,
+      "step": 26626
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.374436855316162,
+      "learning_rate": 4.296404969600846e-05,
+      "loss": 1.6097,
+      "step": 26627
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.073059558868408,
+      "learning_rate": 4.2963785355537935e-05,
+      "loss": 1.3712,
+      "step": 26628
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.061960458755493,
+      "learning_rate": 4.296352101506741e-05,
+      "loss": 1.4571,
+      "step": 26629
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.5237133502960205,
+      "learning_rate": 4.296325667459688e-05,
+      "loss": 1.129,
+      "step": 26630
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.420976400375366,
+      "learning_rate": 4.2962992334126356e-05,
+      "loss": 1.553,
+      "step": 26631
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8532462120056152,
+      "learning_rate": 4.296272799365583e-05,
+      "loss": 1.1858,
+      "step": 26632
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9009064435958862,
+      "learning_rate": 4.29624636531853e-05,
+      "loss": 1.3763,
+      "step": 26633
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1291298866271973,
+      "learning_rate": 4.2962199312714777e-05,
+      "loss": 1.3644,
+      "step": 26634
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8805956840515137,
+      "learning_rate": 4.296193497224425e-05,
+      "loss": 1.2673,
+      "step": 26635
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0535378456115723,
+      "learning_rate": 4.2961670631773724e-05,
+      "loss": 1.2795,
+      "step": 26636
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7740556001663208,
+      "learning_rate": 4.2961406291303204e-05,
+      "loss": 1.2758,
+      "step": 26637
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6534473896026611,
+      "learning_rate": 4.296114195083267e-05,
+      "loss": 1.4088,
+      "step": 26638
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.928244948387146,
+      "learning_rate": 4.296087761036215e-05,
+      "loss": 1.3516,
+      "step": 26639
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8651834726333618,
+      "learning_rate": 4.296061326989162e-05,
+      "loss": 1.2638,
+      "step": 26640
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.262465238571167,
+      "learning_rate": 4.29603489294211e-05,
+      "loss": 1.3621,
+      "step": 26641
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.063854694366455,
+      "learning_rate": 4.296008458895057e-05,
+      "loss": 1.3584,
+      "step": 26642
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.724142074584961,
+      "learning_rate": 4.2959820248480045e-05,
+      "loss": 2.2921,
+      "step": 26643
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.932176113128662,
+      "learning_rate": 4.295955590800952e-05,
+      "loss": 1.226,
+      "step": 26644
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9684901237487793,
+      "learning_rate": 4.295929156753899e-05,
+      "loss": 1.1121,
+      "step": 26645
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.696055293083191,
+      "learning_rate": 4.2959027227068466e-05,
+      "loss": 1.081,
+      "step": 26646
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9440646171569824,
+      "learning_rate": 4.295876288659794e-05,
+      "loss": 1.4583,
+      "step": 26647
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.487623453140259,
+      "learning_rate": 4.295849854612741e-05,
+      "loss": 1.3334,
+      "step": 26648
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.657069206237793,
+      "learning_rate": 4.295823420565689e-05,
+      "loss": 1.527,
+      "step": 26649
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.431333303451538,
+      "learning_rate": 4.295796986518636e-05,
+      "loss": 1.4566,
+      "step": 26650
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.247295618057251,
+      "learning_rate": 4.2957705524715834e-05,
+      "loss": 1.3923,
+      "step": 26651
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9316853284835815,
+      "learning_rate": 4.2957441184245314e-05,
+      "loss": 1.3587,
+      "step": 26652
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1245763301849365,
+      "learning_rate": 4.295717684377478e-05,
+      "loss": 1.3352,
+      "step": 26653
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7062478065490723,
+      "learning_rate": 4.295691250330426e-05,
+      "loss": 2.0171,
+      "step": 26654
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.742242693901062,
+      "learning_rate": 4.295664816283373e-05,
+      "loss": 1.289,
+      "step": 26655
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.400498628616333,
+      "learning_rate": 4.295638382236321e-05,
+      "loss": 1.3575,
+      "step": 26656
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2174885272979736,
+      "learning_rate": 4.295611948189268e-05,
+      "loss": 1.3123,
+      "step": 26657
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.3464014530181885,
+      "learning_rate": 4.295585514142215e-05,
+      "loss": 1.3589,
+      "step": 26658
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0146336555480957,
+      "learning_rate": 4.295559080095163e-05,
+      "loss": 1.3418,
+      "step": 26659
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.236361503601074,
+      "learning_rate": 4.2955326460481096e-05,
+      "loss": 1.9762,
+      "step": 26660
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7662014961242676,
+      "learning_rate": 4.2955062120010576e-05,
+      "loss": 1.2757,
+      "step": 26661
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.6727449893951416,
+      "learning_rate": 4.295479777954005e-05,
+      "loss": 1.3631,
+      "step": 26662
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.614408493041992,
+      "learning_rate": 4.2954533439069524e-05,
+      "loss": 1.2878,
+      "step": 26663
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.944029688835144,
+      "learning_rate": 4.2954269098599e-05,
+      "loss": 1.214,
+      "step": 26664
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.3315937519073486,
+      "learning_rate": 4.295400475812847e-05,
+      "loss": 1.2269,
+      "step": 26665
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.5893363952636719,
+      "learning_rate": 4.2953740417657944e-05,
+      "loss": 1.1701,
+      "step": 26666
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.970716714859009,
+      "learning_rate": 4.295347607718742e-05,
+      "loss": 1.3544,
+      "step": 26667
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9268817901611328,
+      "learning_rate": 4.295321173671689e-05,
+      "loss": 1.4127,
+      "step": 26668
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9556313753128052,
+      "learning_rate": 4.2952947396246365e-05,
+      "loss": 1.3231,
+      "step": 26669
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0995912551879883,
+      "learning_rate": 4.295268305577584e-05,
+      "loss": 1.2746,
+      "step": 26670
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.00044846534729,
+      "learning_rate": 4.295241871530531e-05,
+      "loss": 1.4096,
+      "step": 26671
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0529816150665283,
+      "learning_rate": 4.295215437483479e-05,
+      "loss": 1.3679,
+      "step": 26672
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2765965461730957,
+      "learning_rate": 4.295189003436426e-05,
+      "loss": 1.3072,
+      "step": 26673
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.3784289360046387,
+      "learning_rate": 4.295162569389374e-05,
+      "loss": 1.3883,
+      "step": 26674
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.3283164501190186,
+      "learning_rate": 4.2951361353423206e-05,
+      "loss": 1.4232,
+      "step": 26675
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8981870412826538,
+      "learning_rate": 4.295109701295269e-05,
+      "loss": 1.2325,
+      "step": 26676
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.118948221206665,
+      "learning_rate": 4.295083267248216e-05,
+      "loss": 1.3991,
+      "step": 26677
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9941354990005493,
+      "learning_rate": 4.2950568332011634e-05,
+      "loss": 1.2792,
+      "step": 26678
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.9027090072631836,
+      "learning_rate": 4.295030399154111e-05,
+      "loss": 1.4612,
+      "step": 26679
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8239461183547974,
+      "learning_rate": 4.295003965107058e-05,
+      "loss": 1.4324,
+      "step": 26680
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.425577402114868,
+      "learning_rate": 4.2949775310600055e-05,
+      "loss": 1.4145,
+      "step": 26681
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.002106189727783,
+      "learning_rate": 4.294951097012953e-05,
+      "loss": 1.215,
+      "step": 26682
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.743581533432007,
+      "learning_rate": 4.2949246629659e-05,
+      "loss": 1.3951,
+      "step": 26683
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.350241184234619,
+      "learning_rate": 4.2948982289188475e-05,
+      "loss": 1.3554,
+      "step": 26684
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8757524490356445,
+      "learning_rate": 4.294871794871795e-05,
+      "loss": 1.2414,
+      "step": 26685
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.785209059715271,
+      "learning_rate": 4.294845360824742e-05,
+      "loss": 1.3085,
+      "step": 26686
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1881730556488037,
+      "learning_rate": 4.29481892677769e-05,
+      "loss": 1.2163,
+      "step": 26687
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8245905637741089,
+      "learning_rate": 4.294792492730637e-05,
+      "loss": 1.1589,
+      "step": 26688
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.536569833755493,
+      "learning_rate": 4.294766058683585e-05,
+      "loss": 1.4709,
+      "step": 26689
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9763234853744507,
+      "learning_rate": 4.294739624636532e-05,
+      "loss": 1.3725,
+      "step": 26690
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1373939514160156,
+      "learning_rate": 4.29471319058948e-05,
+      "loss": 1.3401,
+      "step": 26691
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.934729814529419,
+      "learning_rate": 4.294686756542427e-05,
+      "loss": 1.208,
+      "step": 26692
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.167100667953491,
+      "learning_rate": 4.294660322495374e-05,
+      "loss": 1.2678,
+      "step": 26693
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8472574949264526,
+      "learning_rate": 4.294633888448322e-05,
+      "loss": 1.3625,
+      "step": 26694
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.6608145236968994,
+      "learning_rate": 4.2946074544012685e-05,
+      "loss": 1.3387,
+      "step": 26695
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8397291898727417,
+      "learning_rate": 4.2945810203542165e-05,
+      "loss": 1.3717,
+      "step": 26696
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6947077512741089,
+      "learning_rate": 4.294554586307164e-05,
+      "loss": 2.1501,
+      "step": 26697
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0145301818847656,
+      "learning_rate": 4.294528152260111e-05,
+      "loss": 1.2831,
+      "step": 26698
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6935360431671143,
+      "learning_rate": 4.2945017182130586e-05,
+      "loss": 1.7852,
+      "step": 26699
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.4933576583862305,
+      "learning_rate": 4.294475284166006e-05,
+      "loss": 1.5446,
+      "step": 26700
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.358511447906494,
+      "learning_rate": 4.294448850118953e-05,
+      "loss": 1.2569,
+      "step": 26701
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8282514810562134,
+      "learning_rate": 4.294422416071901e-05,
+      "loss": 1.3566,
+      "step": 26702
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9659799337387085,
+      "learning_rate": 4.294395982024848e-05,
+      "loss": 1.2653,
+      "step": 26703
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.731680393218994,
+      "learning_rate": 4.2943695479777954e-05,
+      "loss": 1.402,
+      "step": 26704
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6049778461456299,
+      "learning_rate": 4.294343113930743e-05,
+      "loss": 1.2879,
+      "step": 26705
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.4381651878356934,
+      "learning_rate": 4.29431667988369e-05,
+      "loss": 1.4215,
+      "step": 26706
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9222750663757324,
+      "learning_rate": 4.294290245836638e-05,
+      "loss": 0.9911,
+      "step": 26707
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.5824954509735107,
+      "learning_rate": 4.294263811789585e-05,
+      "loss": 1.1358,
+      "step": 26708
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8063838481903076,
+      "learning_rate": 4.294237377742533e-05,
+      "loss": 1.1935,
+      "step": 26709
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1098785400390625,
+      "learning_rate": 4.2942109436954795e-05,
+      "loss": 1.4182,
+      "step": 26710
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7726656198501587,
+      "learning_rate": 4.2941845096484275e-05,
+      "loss": 1.7355,
+      "step": 26711
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.367921829223633,
+      "learning_rate": 4.294158075601375e-05,
+      "loss": 1.4594,
+      "step": 26712
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0050830841064453,
+      "learning_rate": 4.294131641554322e-05,
+      "loss": 1.3817,
+      "step": 26713
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2773566246032715,
+      "learning_rate": 4.2941052075072696e-05,
+      "loss": 1.5273,
+      "step": 26714
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7359285354614258,
+      "learning_rate": 4.294078773460217e-05,
+      "loss": 3.7786,
+      "step": 26715
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8037822246551514,
+      "learning_rate": 4.294052339413164e-05,
+      "loss": 1.2228,
+      "step": 26716
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.036712884902954,
+      "learning_rate": 4.294025905366112e-05,
+      "loss": 1.3466,
+      "step": 26717
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.018578290939331,
+      "learning_rate": 4.293999471319059e-05,
+      "loss": 1.2144,
+      "step": 26718
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.173433303833008,
+      "learning_rate": 4.2939730372720064e-05,
+      "loss": 1.3677,
+      "step": 26719
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.6771860122680664,
+      "learning_rate": 4.293946603224954e-05,
+      "loss": 1.4843,
+      "step": 26720
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7280642986297607,
+      "learning_rate": 4.293920169177901e-05,
+      "loss": 1.2171,
+      "step": 26721
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8292521238327026,
+      "learning_rate": 4.293893735130849e-05,
+      "loss": 1.3585,
+      "step": 26722
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.7259624004364014,
+      "learning_rate": 4.293867301083796e-05,
+      "loss": 1.3571,
+      "step": 26723
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6038789749145508,
+      "learning_rate": 4.293840867036744e-05,
+      "loss": 1.3384,
+      "step": 26724
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8885118961334229,
+      "learning_rate": 4.2938144329896905e-05,
+      "loss": 1.2932,
+      "step": 26725
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.636431932449341,
+      "learning_rate": 4.2937879989426386e-05,
+      "loss": 1.4536,
+      "step": 26726
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6613829135894775,
+      "learning_rate": 4.293761564895586e-05,
+      "loss": 1.3041,
+      "step": 26727
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.096306562423706,
+      "learning_rate": 4.293735130848533e-05,
+      "loss": 1.283,
+      "step": 26728
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.523066520690918,
+      "learning_rate": 4.2937086968014806e-05,
+      "loss": 1.3938,
+      "step": 26729
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0816457271575928,
+      "learning_rate": 4.293682262754427e-05,
+      "loss": 1.2849,
+      "step": 26730
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1034066677093506,
+      "learning_rate": 4.2936558287073754e-05,
+      "loss": 1.2755,
+      "step": 26731
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.112583875656128,
+      "learning_rate": 4.293629394660323e-05,
+      "loss": 1.3394,
+      "step": 26732
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.867667555809021,
+      "learning_rate": 4.29360296061327e-05,
+      "loss": 1.1926,
+      "step": 26733
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2297890186309814,
+      "learning_rate": 4.2935765265662174e-05,
+      "loss": 1.3044,
+      "step": 26734
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.8055882453918457,
+      "learning_rate": 4.293550092519165e-05,
+      "loss": 1.354,
+      "step": 26735
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.371732473373413,
+      "learning_rate": 4.293523658472112e-05,
+      "loss": 1.403,
+      "step": 26736
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7381947040557861,
+      "learning_rate": 4.29349722442506e-05,
+      "loss": 1.1469,
+      "step": 26737
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.111713409423828,
+      "learning_rate": 4.293470790378007e-05,
+      "loss": 1.2952,
+      "step": 26738
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8317867517471313,
+      "learning_rate": 4.293444356330954e-05,
+      "loss": 1.3182,
+      "step": 26739
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.412005662918091,
+      "learning_rate": 4.2934179222839016e-05,
+      "loss": 1.3688,
+      "step": 26740
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0278236865997314,
+      "learning_rate": 4.293391488236849e-05,
+      "loss": 1.2307,
+      "step": 26741
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.309170961380005,
+      "learning_rate": 4.293365054189797e-05,
+      "loss": 1.2219,
+      "step": 26742
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.592885971069336,
+      "learning_rate": 4.2933386201427436e-05,
+      "loss": 1.5811,
+      "step": 26743
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.18469500541687,
+      "learning_rate": 4.293312186095692e-05,
+      "loss": 1.312,
+      "step": 26744
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8908287286758423,
+      "learning_rate": 4.2932857520486384e-05,
+      "loss": 1.3279,
+      "step": 26745
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.197901487350464,
+      "learning_rate": 4.2932593180015864e-05,
+      "loss": 1.2315,
+      "step": 26746
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.770775318145752,
+      "learning_rate": 4.293232883954534e-05,
+      "loss": 1.2939,
+      "step": 26747
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2084896564483643,
+      "learning_rate": 4.293206449907481e-05,
+      "loss": 1.3735,
+      "step": 26748
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.6248040199279785,
+      "learning_rate": 4.2931800158604285e-05,
+      "loss": 1.5723,
+      "step": 26749
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.1364126205444336,
+      "learning_rate": 4.293153581813376e-05,
+      "loss": 1.4515,
+      "step": 26750
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.317772388458252,
+      "learning_rate": 4.293127147766323e-05,
+      "loss": 1.3081,
+      "step": 26751
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.244814872741699,
+      "learning_rate": 4.2931007137192705e-05,
+      "loss": 1.4365,
+      "step": 26752
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9947282075881958,
+      "learning_rate": 4.293074279672218e-05,
+      "loss": 1.2378,
+      "step": 26753
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8200079202651978,
+      "learning_rate": 4.293047845625165e-05,
+      "loss": 1.2939,
+      "step": 26754
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7383370399475098,
+      "learning_rate": 4.2930214115781126e-05,
+      "loss": 1.2551,
+      "step": 26755
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9553065299987793,
+      "learning_rate": 4.29299497753106e-05,
+      "loss": 1.2425,
+      "step": 26756
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8795864582061768,
+      "learning_rate": 4.292968543484008e-05,
+      "loss": 1.3505,
+      "step": 26757
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9289506673812866,
+      "learning_rate": 4.292942109436955e-05,
+      "loss": 1.2836,
+      "step": 26758
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.3879799842834473,
+      "learning_rate": 4.292915675389903e-05,
+      "loss": 1.4814,
+      "step": 26759
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6498967409133911,
+      "learning_rate": 4.2928892413428494e-05,
+      "loss": 1.2743,
+      "step": 26760
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.062398672103882,
+      "learning_rate": 4.2928628072957974e-05,
+      "loss": 1.1237,
+      "step": 26761
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8881349563598633,
+      "learning_rate": 4.292836373248745e-05,
+      "loss": 1.3864,
+      "step": 26762
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.33072566986084,
+      "learning_rate": 4.292809939201692e-05,
+      "loss": 1.2204,
+      "step": 26763
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.89662504196167,
+      "learning_rate": 4.2927835051546395e-05,
+      "loss": 1.1965,
+      "step": 26764
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7470738887786865,
+      "learning_rate": 4.292757071107587e-05,
+      "loss": 1.2249,
+      "step": 26765
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.012289524078369,
+      "learning_rate": 4.292730637060534e-05,
+      "loss": 1.2949,
+      "step": 26766
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2348155975341797,
+      "learning_rate": 4.2927042030134816e-05,
+      "loss": 1.3482,
+      "step": 26767
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3369593620300293,
+      "learning_rate": 4.292677768966429e-05,
+      "loss": 1.4513,
+      "step": 26768
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7155776023864746,
+      "learning_rate": 4.292651334919376e-05,
+      "loss": 1.3143,
+      "step": 26769
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0833804607391357,
+      "learning_rate": 4.2926249008723236e-05,
+      "loss": 1.3082,
+      "step": 26770
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.454301595687866,
+      "learning_rate": 4.292598466825271e-05,
+      "loss": 1.3953,
+      "step": 26771
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3718502521514893,
+      "learning_rate": 4.292572032778219e-05,
+      "loss": 1.3361,
+      "step": 26772
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.5799639225006104,
+      "learning_rate": 4.292545598731166e-05,
+      "loss": 1.4253,
+      "step": 26773
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3846442699432373,
+      "learning_rate": 4.292519164684114e-05,
+      "loss": 1.4062,
+      "step": 26774
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.809699535369873,
+      "learning_rate": 4.2924927306370604e-05,
+      "loss": 1.4607,
+      "step": 26775
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.701494812965393,
+      "learning_rate": 4.292466296590008e-05,
+      "loss": 1.2146,
+      "step": 26776
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.054879665374756,
+      "learning_rate": 4.292439862542956e-05,
+      "loss": 1.2688,
+      "step": 26777
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8734883069992065,
+      "learning_rate": 4.2924134284959025e-05,
+      "loss": 1.2833,
+      "step": 26778
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9565967321395874,
+      "learning_rate": 4.2923869944488505e-05,
+      "loss": 1.2568,
+      "step": 26779
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.430373191833496,
+      "learning_rate": 4.292360560401797e-05,
+      "loss": 1.0877,
+      "step": 26780
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.4250082969665527,
+      "learning_rate": 4.292334126354745e-05,
+      "loss": 1.3559,
+      "step": 26781
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.210132360458374,
+      "learning_rate": 4.2923076923076926e-05,
+      "loss": 1.2004,
+      "step": 26782
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7256451845169067,
+      "learning_rate": 4.29228125826064e-05,
+      "loss": 1.136,
+      "step": 26783
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1419849395751953,
+      "learning_rate": 4.292254824213587e-05,
+      "loss": 1.36,
+      "step": 26784
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7078397274017334,
+      "learning_rate": 4.292228390166535e-05,
+      "loss": 1.2421,
+      "step": 26785
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.4479520320892334,
+      "learning_rate": 4.292201956119482e-05,
+      "loss": 1.3051,
+      "step": 26786
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.8572964668273926,
+      "learning_rate": 4.2921755220724294e-05,
+      "loss": 1.5238,
+      "step": 26787
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.094904661178589,
+      "learning_rate": 4.292149088025377e-05,
+      "loss": 1.3523,
+      "step": 26788
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.304270029067993,
+      "learning_rate": 4.292122653978324e-05,
+      "loss": 1.3951,
+      "step": 26789
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7072731256484985,
+      "learning_rate": 4.2920962199312715e-05,
+      "loss": 1.5753,
+      "step": 26790
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3398311138153076,
+      "learning_rate": 4.292069785884219e-05,
+      "loss": 1.3945,
+      "step": 26791
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.158578634262085,
+      "learning_rate": 4.292043351837167e-05,
+      "loss": 1.2527,
+      "step": 26792
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0236103534698486,
+      "learning_rate": 4.2920169177901135e-05,
+      "loss": 1.2155,
+      "step": 26793
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8738548755645752,
+      "learning_rate": 4.2919904837430616e-05,
+      "loss": 1.3285,
+      "step": 26794
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7404783964157104,
+      "learning_rate": 4.291964049696008e-05,
+      "loss": 1.2417,
+      "step": 26795
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0820775032043457,
+      "learning_rate": 4.291937615648956e-05,
+      "loss": 1.2389,
+      "step": 26796
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7854431867599487,
+      "learning_rate": 4.2919111816019036e-05,
+      "loss": 1.3012,
+      "step": 26797
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.377126693725586,
+      "learning_rate": 4.291884747554851e-05,
+      "loss": 1.3719,
+      "step": 26798
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.2391133308410645,
+      "learning_rate": 4.2918583135077983e-05,
+      "loss": 1.2838,
+      "step": 26799
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3154733180999756,
+      "learning_rate": 4.291831879460746e-05,
+      "loss": 1.4349,
+      "step": 26800
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.478367805480957,
+      "learning_rate": 4.291805445413693e-05,
+      "loss": 1.3554,
+      "step": 26801
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8100371360778809,
+      "learning_rate": 4.2917790113666404e-05,
+      "loss": 1.3906,
+      "step": 26802
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9964174032211304,
+      "learning_rate": 4.291752577319588e-05,
+      "loss": 1.228,
+      "step": 26803
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1202306747436523,
+      "learning_rate": 4.291726143272535e-05,
+      "loss": 1.2059,
+      "step": 26804
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9999072551727295,
+      "learning_rate": 4.2916997092254825e-05,
+      "loss": 1.7644,
+      "step": 26805
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0841023921966553,
+      "learning_rate": 4.29167327517843e-05,
+      "loss": 1.3876,
+      "step": 26806
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0829761028289795,
+      "learning_rate": 4.291646841131378e-05,
+      "loss": 1.3496,
+      "step": 26807
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.260584592819214,
+      "learning_rate": 4.2916204070843246e-05,
+      "loss": 1.4067,
+      "step": 26808
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0895278453826904,
+      "learning_rate": 4.2915939730372726e-05,
+      "loss": 1.3704,
+      "step": 26809
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8259648084640503,
+      "learning_rate": 4.291567538990219e-05,
+      "loss": 1.2939,
+      "step": 26810
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3386900424957275,
+      "learning_rate": 4.2915411049431666e-05,
+      "loss": 1.3105,
+      "step": 26811
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.146817207336426,
+      "learning_rate": 4.291514670896115e-05,
+      "loss": 1.323,
+      "step": 26812
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8492931127548218,
+      "learning_rate": 4.2914882368490613e-05,
+      "loss": 1.2711,
+      "step": 26813
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.281365156173706,
+      "learning_rate": 4.2914618028020094e-05,
+      "loss": 1.3933,
+      "step": 26814
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.930929660797119,
+      "learning_rate": 4.291435368754956e-05,
+      "loss": 1.4581,
+      "step": 26815
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9703645706176758,
+      "learning_rate": 4.291408934707904e-05,
+      "loss": 1.3219,
+      "step": 26816
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1956818103790283,
+      "learning_rate": 4.2913825006608515e-05,
+      "loss": 1.353,
+      "step": 26817
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8884772062301636,
+      "learning_rate": 4.291356066613799e-05,
+      "loss": 1.2654,
+      "step": 26818
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.742265224456787,
+      "learning_rate": 4.291329632566746e-05,
+      "loss": 1.4871,
+      "step": 26819
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9788281917572021,
+      "learning_rate": 4.2913031985196935e-05,
+      "loss": 1.3751,
+      "step": 26820
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7289663553237915,
+      "learning_rate": 4.291276764472641e-05,
+      "loss": 1.2032,
+      "step": 26821
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.2548248767852783,
+      "learning_rate": 4.291250330425588e-05,
+      "loss": 1.4676,
+      "step": 26822
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.129317045211792,
+      "learning_rate": 4.2912238963785356e-05,
+      "loss": 2.0704,
+      "step": 26823
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9442373514175415,
+      "learning_rate": 4.291197462331483e-05,
+      "loss": 1.1135,
+      "step": 26824
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.083383321762085,
+      "learning_rate": 4.29117102828443e-05,
+      "loss": 1.3693,
+      "step": 26825
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.17673659324646,
+      "learning_rate": 4.291144594237378e-05,
+      "loss": 1.2881,
+      "step": 26826
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3696491718292236,
+      "learning_rate": 4.291118160190326e-05,
+      "loss": 1.4336,
+      "step": 26827
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1995999813079834,
+      "learning_rate": 4.2910917261432724e-05,
+      "loss": 1.408,
+      "step": 26828
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.8091914653778076,
+      "learning_rate": 4.2910652920962204e-05,
+      "loss": 1.3784,
+      "step": 26829
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.926193118095398,
+      "learning_rate": 4.291038858049167e-05,
+      "loss": 1.2528,
+      "step": 26830
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.851860523223877,
+      "learning_rate": 4.291012424002115e-05,
+      "loss": 1.2581,
+      "step": 26831
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.484008550643921,
+      "learning_rate": 4.2909859899550625e-05,
+      "loss": 1.3382,
+      "step": 26832
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9272700548171997,
+      "learning_rate": 4.29095955590801e-05,
+      "loss": 1.1236,
+      "step": 26833
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.627955675125122,
+      "learning_rate": 4.290933121860957e-05,
+      "loss": 1.1822,
+      "step": 26834
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9988832473754883,
+      "learning_rate": 4.2909066878139046e-05,
+      "loss": 1.7247,
+      "step": 26835
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8949447870254517,
+      "learning_rate": 4.290880253766852e-05,
+      "loss": 1.2764,
+      "step": 26836
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.079134464263916,
+      "learning_rate": 4.290853819719799e-05,
+      "loss": 1.2662,
+      "step": 26837
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0900862216949463,
+      "learning_rate": 4.2908273856727466e-05,
+      "loss": 1.2458,
+      "step": 26838
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.969027042388916,
+      "learning_rate": 4.290800951625694e-05,
+      "loss": 1.2459,
+      "step": 26839
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.360776662826538,
+      "learning_rate": 4.2907745175786413e-05,
+      "loss": 2.6095,
+      "step": 26840
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9886088371276855,
+      "learning_rate": 4.290748083531589e-05,
+      "loss": 1.2617,
+      "step": 26841
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.134097099304199,
+      "learning_rate": 4.290721649484537e-05,
+      "loss": 1.3373,
+      "step": 26842
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0275521278381348,
+      "learning_rate": 4.2906952154374834e-05,
+      "loss": 1.2232,
+      "step": 26843
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8200347423553467,
+      "learning_rate": 4.2906687813904314e-05,
+      "loss": 1.1295,
+      "step": 26844
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3698019981384277,
+      "learning_rate": 4.290642347343378e-05,
+      "loss": 1.5406,
+      "step": 26845
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.989207148551941,
+      "learning_rate": 4.290615913296326e-05,
+      "loss": 1.3571,
+      "step": 26846
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1744565963745117,
+      "learning_rate": 4.2905894792492735e-05,
+      "loss": 1.4342,
+      "step": 26847
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8733090162277222,
+      "learning_rate": 4.29056304520222e-05,
+      "loss": 1.3088,
+      "step": 26848
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.68910551071167,
+      "learning_rate": 4.290536611155168e-05,
+      "loss": 1.3307,
+      "step": 26849
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.731896996498108,
+      "learning_rate": 4.290510177108115e-05,
+      "loss": 1.3558,
+      "step": 26850
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6856657266616821,
+      "learning_rate": 4.290483743061063e-05,
+      "loss": 1.1945,
+      "step": 26851
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0116212368011475,
+      "learning_rate": 4.29045730901401e-05,
+      "loss": 1.3506,
+      "step": 26852
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.518852472305298,
+      "learning_rate": 4.290430874966958e-05,
+      "loss": 1.3896,
+      "step": 26853
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.5340709686279297,
+      "learning_rate": 4.290404440919905e-05,
+      "loss": 1.5937,
+      "step": 26854
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6436094045639038,
+      "learning_rate": 4.2903780068728524e-05,
+      "loss": 1.2671,
+      "step": 26855
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9386918544769287,
+      "learning_rate": 4.2903515728258e-05,
+      "loss": 1.39,
+      "step": 26856
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8176653385162354,
+      "learning_rate": 4.290325138778747e-05,
+      "loss": 1.7159,
+      "step": 26857
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8533248901367188,
+      "learning_rate": 4.2902987047316944e-05,
+      "loss": 1.2593,
+      "step": 26858
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.740063190460205,
+      "learning_rate": 4.290272270684642e-05,
+      "loss": 1.2114,
+      "step": 26859
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8126662969589233,
+      "learning_rate": 4.290245836637589e-05,
+      "loss": 1.3261,
+      "step": 26860
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.722365379333496,
+      "learning_rate": 4.2902194025905365e-05,
+      "loss": 1.2566,
+      "step": 26861
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3065478801727295,
+      "learning_rate": 4.2901929685434846e-05,
+      "loss": 1.3938,
+      "step": 26862
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.4618959426879883,
+      "learning_rate": 4.290166534496431e-05,
+      "loss": 1.0664,
+      "step": 26863
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3401095867156982,
+      "learning_rate": 4.290140100449379e-05,
+      "loss": 1.3059,
+      "step": 26864
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6522444486618042,
+      "learning_rate": 4.290113666402326e-05,
+      "loss": 1.1943,
+      "step": 26865
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.007765531539917,
+      "learning_rate": 4.290087232355274e-05,
+      "loss": 1.4121,
+      "step": 26866
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6420093774795532,
+      "learning_rate": 4.2900607983082213e-05,
+      "loss": 1.3466,
+      "step": 26867
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.076979160308838,
+      "learning_rate": 4.290034364261169e-05,
+      "loss": 1.3296,
+      "step": 26868
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.726222276687622,
+      "learning_rate": 4.290007930214116e-05,
+      "loss": 1.183,
+      "step": 26869
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9110496044158936,
+      "learning_rate": 4.2899814961670634e-05,
+      "loss": 1.2378,
+      "step": 26870
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.482128143310547,
+      "learning_rate": 4.289955062120011e-05,
+      "loss": 1.4894,
+      "step": 26871
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.15618634223938,
+      "learning_rate": 4.289928628072958e-05,
+      "loss": 1.4393,
+      "step": 26872
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9860665798187256,
+      "learning_rate": 4.2899021940259055e-05,
+      "loss": 1.3524,
+      "step": 26873
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9153872728347778,
+      "learning_rate": 4.289875759978853e-05,
+      "loss": 1.221,
+      "step": 26874
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.913231372833252,
+      "learning_rate": 4.2898493259318e-05,
+      "loss": 1.3237,
+      "step": 26875
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.369640588760376,
+      "learning_rate": 4.2898228918847476e-05,
+      "loss": 1.4286,
+      "step": 26876
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.352849006652832,
+      "learning_rate": 4.2897964578376956e-05,
+      "loss": 1.271,
+      "step": 26877
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.088578462600708,
+      "learning_rate": 4.289770023790642e-05,
+      "loss": 1.4849,
+      "step": 26878
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.908557176589966,
+      "learning_rate": 4.28974358974359e-05,
+      "loss": 1.6318,
+      "step": 26879
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8712639808654785,
+      "learning_rate": 4.289717155696537e-05,
+      "loss": 1.3089,
+      "step": 26880
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9558837413787842,
+      "learning_rate": 4.289690721649485e-05,
+      "loss": 1.3299,
+      "step": 26881
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.011449098587036,
+      "learning_rate": 4.2896642876024324e-05,
+      "loss": 1.2538,
+      "step": 26882
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.015662431716919,
+      "learning_rate": 4.28963785355538e-05,
+      "loss": 1.1326,
+      "step": 26883
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.376030683517456,
+      "learning_rate": 4.289611419508327e-05,
+      "loss": 1.2978,
+      "step": 26884
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.465733528137207,
+      "learning_rate": 4.289584985461274e-05,
+      "loss": 2.5345,
+      "step": 26885
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0356459617614746,
+      "learning_rate": 4.289558551414222e-05,
+      "loss": 1.3696,
+      "step": 26886
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8519322872161865,
+      "learning_rate": 4.289532117367169e-05,
+      "loss": 1.2107,
+      "step": 26887
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8989681005477905,
+      "learning_rate": 4.2895056833201165e-05,
+      "loss": 1.2994,
+      "step": 26888
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.082314968109131,
+      "learning_rate": 4.289479249273064e-05,
+      "loss": 1.3616,
+      "step": 26889
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.670811414718628,
+      "learning_rate": 4.289452815226011e-05,
+      "loss": 1.3634,
+      "step": 26890
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.0146682262420654,
+      "learning_rate": 4.2894263811789586e-05,
+      "loss": 1.2763,
+      "step": 26891
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.098642349243164,
+      "learning_rate": 4.2893999471319066e-05,
+      "loss": 2.2433,
+      "step": 26892
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.312032699584961,
+      "learning_rate": 4.289373513084853e-05,
+      "loss": 1.4246,
+      "step": 26893
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0505850315093994,
+      "learning_rate": 4.2893470790378007e-05,
+      "loss": 1.3313,
+      "step": 26894
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.69125235080719,
+      "learning_rate": 4.289320644990748e-05,
+      "loss": 1.2041,
+      "step": 26895
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.863808512687683,
+      "learning_rate": 4.2892942109436954e-05,
+      "loss": 2.5053,
+      "step": 26896
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.450629949569702,
+      "learning_rate": 4.2892677768966434e-05,
+      "loss": 1.3147,
+      "step": 26897
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.2817158699035645,
+      "learning_rate": 4.28924134284959e-05,
+      "loss": 1.4419,
+      "step": 26898
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.61741304397583,
+      "learning_rate": 4.289214908802538e-05,
+      "loss": 1.4079,
+      "step": 26899
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.4110374450683594,
+      "learning_rate": 4.289188474755485e-05,
+      "loss": 1.4155,
+      "step": 26900
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8864476680755615,
+      "learning_rate": 4.289162040708433e-05,
+      "loss": 1.3378,
+      "step": 26901
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8122221231460571,
+      "learning_rate": 4.28913560666138e-05,
+      "loss": 1.56,
+      "step": 26902
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.4097564220428467,
+      "learning_rate": 4.2891091726143276e-05,
+      "loss": 1.172,
+      "step": 26903
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0662436485290527,
+      "learning_rate": 4.289082738567275e-05,
+      "loss": 1.2103,
+      "step": 26904
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1545896530151367,
+      "learning_rate": 4.289056304520222e-05,
+      "loss": 1.2307,
+      "step": 26905
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.900732398033142,
+      "learning_rate": 4.2890298704731696e-05,
+      "loss": 1.2516,
+      "step": 26906
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.834173560142517,
+      "learning_rate": 4.289003436426117e-05,
+      "loss": 1.2366,
+      "step": 26907
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.2106399536132812,
+      "learning_rate": 4.288977002379064e-05,
+      "loss": 1.214,
+      "step": 26908
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7292383909225464,
+      "learning_rate": 4.288950568332012e-05,
+      "loss": 1.1558,
+      "step": 26909
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7701562643051147,
+      "learning_rate": 4.288924134284959e-05,
+      "loss": 1.1712,
+      "step": 26910
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0434646606445312,
+      "learning_rate": 4.2888977002379064e-05,
+      "loss": 1.1787,
+      "step": 26911
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.733206748962402,
+      "learning_rate": 4.2888712661908544e-05,
+      "loss": 1.34,
+      "step": 26912
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.224317789077759,
+      "learning_rate": 4.288844832143801e-05,
+      "loss": 1.4931,
+      "step": 26913
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9684216976165771,
+      "learning_rate": 4.288818398096749e-05,
+      "loss": 1.4213,
+      "step": 26914
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.974783182144165,
+      "learning_rate": 4.288791964049696e-05,
+      "loss": 1.2116,
+      "step": 26915
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8821394443511963,
+      "learning_rate": 4.288765530002644e-05,
+      "loss": 1.354,
+      "step": 26916
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.051628589630127,
+      "learning_rate": 4.288739095955591e-05,
+      "loss": 1.1119,
+      "step": 26917
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.4495794773101807,
+      "learning_rate": 4.2887126619085386e-05,
+      "loss": 1.4626,
+      "step": 26918
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0822675228118896,
+      "learning_rate": 4.288686227861486e-05,
+      "loss": 1.2757,
+      "step": 26919
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0265402793884277,
+      "learning_rate": 4.2886597938144326e-05,
+      "loss": 1.2377,
+      "step": 26920
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6397041082382202,
+      "learning_rate": 4.2886333597673807e-05,
+      "loss": 1.0799,
+      "step": 26921
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0365893840789795,
+      "learning_rate": 4.288606925720328e-05,
+      "loss": 1.2886,
+      "step": 26922
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.2700204849243164,
+      "learning_rate": 4.2885804916732754e-05,
+      "loss": 1.3218,
+      "step": 26923
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1994946002960205,
+      "learning_rate": 4.288554057626223e-05,
+      "loss": 1.3663,
+      "step": 26924
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.125570058822632,
+      "learning_rate": 4.28852762357917e-05,
+      "loss": 1.1967,
+      "step": 26925
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.726181149482727,
+      "learning_rate": 4.2885011895321174e-05,
+      "loss": 1.3527,
+      "step": 26926
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9915188550949097,
+      "learning_rate": 4.2884747554850655e-05,
+      "loss": 1.2732,
+      "step": 26927
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7012403011322021,
+      "learning_rate": 4.288448321438012e-05,
+      "loss": 1.1516,
+      "step": 26928
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.967139959335327,
+      "learning_rate": 4.2884218873909595e-05,
+      "loss": 1.3279,
+      "step": 26929
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1209969520568848,
+      "learning_rate": 4.288395453343907e-05,
+      "loss": 1.3536,
+      "step": 26930
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.374419927597046,
+      "learning_rate": 4.288369019296854e-05,
+      "loss": 1.322,
+      "step": 26931
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.780168056488037,
+      "learning_rate": 4.288342585249802e-05,
+      "loss": 1.2823,
+      "step": 26932
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9513721466064453,
+      "learning_rate": 4.288316151202749e-05,
+      "loss": 1.2885,
+      "step": 26933
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.49839448928833,
+      "learning_rate": 4.288289717155697e-05,
+      "loss": 1.756,
+      "step": 26934
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1570491790771484,
+      "learning_rate": 4.2882632831086437e-05,
+      "loss": 1.2747,
+      "step": 26935
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6290414333343506,
+      "learning_rate": 4.288236849061592e-05,
+      "loss": 1.3147,
+      "step": 26936
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9045332670211792,
+      "learning_rate": 4.288210415014539e-05,
+      "loss": 1.2884,
+      "step": 26937
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.951591968536377,
+      "learning_rate": 4.2881839809674864e-05,
+      "loss": 1.4033,
+      "step": 26938
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8345290422439575,
+      "learning_rate": 4.288157546920434e-05,
+      "loss": 1.3789,
+      "step": 26939
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.9635268449783325,
+      "learning_rate": 4.288131112873381e-05,
+      "loss": 1.2291,
+      "step": 26940
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7958922386169434,
+      "learning_rate": 4.2881046788263285e-05,
+      "loss": 1.3707,
+      "step": 26941
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7888531684875488,
+      "learning_rate": 4.288078244779276e-05,
+      "loss": 1.2386,
+      "step": 26942
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.0211615562438965,
+      "learning_rate": 4.288051810732223e-05,
+      "loss": 1.4467,
+      "step": 26943
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.219085931777954,
+      "learning_rate": 4.2880253766851705e-05,
+      "loss": 1.4543,
+      "step": 26944
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7541602849960327,
+      "learning_rate": 4.287998942638118e-05,
+      "loss": 2.1707,
+      "step": 26945
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.6454601287841797,
+      "learning_rate": 4.287972508591065e-05,
+      "loss": 1.4268,
+      "step": 26946
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.045140027999878,
+      "learning_rate": 4.287946074544013e-05,
+      "loss": 1.4847,
+      "step": 26947
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.6388964653015137,
+      "learning_rate": 4.28791964049696e-05,
+      "loss": 1.2433,
+      "step": 26948
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.4302172660827637,
+      "learning_rate": 4.287893206449908e-05,
+      "loss": 1.4609,
+      "step": 26949
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0824496746063232,
+      "learning_rate": 4.287866772402855e-05,
+      "loss": 1.2474,
+      "step": 26950
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0854666233062744,
+      "learning_rate": 4.287840338355803e-05,
+      "loss": 1.3521,
+      "step": 26951
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.1928019523620605,
+      "learning_rate": 4.28781390430875e-05,
+      "loss": 1.3102,
+      "step": 26952
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.040724515914917,
+      "learning_rate": 4.2877874702616974e-05,
+      "loss": 1.1899,
+      "step": 26953
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.584976315498352,
+      "learning_rate": 4.287761036214645e-05,
+      "loss": 1.1376,
+      "step": 26954
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.755746841430664,
+      "learning_rate": 4.287734602167592e-05,
+      "loss": 1.2026,
+      "step": 26955
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.7220377922058105,
+      "learning_rate": 4.2877081681205395e-05,
+      "loss": 1.6255,
+      "step": 26956
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.081040620803833,
+      "learning_rate": 4.287681734073487e-05,
+      "loss": 1.2226,
+      "step": 26957
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9628337621688843,
+      "learning_rate": 4.287655300026434e-05,
+      "loss": 1.2967,
+      "step": 26958
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7486683130264282,
+      "learning_rate": 4.2876288659793816e-05,
+      "loss": 1.0937,
+      "step": 26959
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.626930832862854,
+      "learning_rate": 4.287602431932329e-05,
+      "loss": 1.2264,
+      "step": 26960
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.4525558948516846,
+      "learning_rate": 4.287575997885276e-05,
+      "loss": 1.0879,
+      "step": 26961
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.801598310470581,
+      "learning_rate": 4.287549563838224e-05,
+      "loss": 1.37,
+      "step": 26962
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9950685501098633,
+      "learning_rate": 4.287523129791171e-05,
+      "loss": 1.3451,
+      "step": 26963
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7756794691085815,
+      "learning_rate": 4.287496695744119e-05,
+      "loss": 1.3356,
+      "step": 26964
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.40574312210083,
+      "learning_rate": 4.287470261697066e-05,
+      "loss": 1.4273,
+      "step": 26965
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.047973871231079,
+      "learning_rate": 4.287443827650013e-05,
+      "loss": 1.3658,
+      "step": 26966
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6264513731002808,
+      "learning_rate": 4.287417393602961e-05,
+      "loss": 1.2005,
+      "step": 26967
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7021715641021729,
+      "learning_rate": 4.287390959555908e-05,
+      "loss": 1.24,
+      "step": 26968
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.2137482166290283,
+      "learning_rate": 4.287364525508856e-05,
+      "loss": 1.3784,
+      "step": 26969
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.857949137687683,
+      "learning_rate": 4.2873380914618025e-05,
+      "loss": 1.2708,
+      "step": 26970
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.117713451385498,
+      "learning_rate": 4.2873116574147505e-05,
+      "loss": 1.1907,
+      "step": 26971
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7664035558700562,
+      "learning_rate": 4.287285223367698e-05,
+      "loss": 1.2374,
+      "step": 26972
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1067442893981934,
+      "learning_rate": 4.287258789320645e-05,
+      "loss": 1.3995,
+      "step": 26973
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8609379529953003,
+      "learning_rate": 4.2872323552735926e-05,
+      "loss": 1.297,
+      "step": 26974
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1915440559387207,
+      "learning_rate": 4.28720592122654e-05,
+      "loss": 1.439,
+      "step": 26975
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.5639944076538086,
+      "learning_rate": 4.287179487179487e-05,
+      "loss": 1.4591,
+      "step": 26976
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3455991744995117,
+      "learning_rate": 4.287153053132435e-05,
+      "loss": 1.3413,
+      "step": 26977
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.7715933322906494,
+      "learning_rate": 4.287126619085382e-05,
+      "loss": 1.4441,
+      "step": 26978
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3416061401367188,
+      "learning_rate": 4.2871001850383294e-05,
+      "loss": 1.3432,
+      "step": 26979
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.61197030544281,
+      "learning_rate": 4.287073750991277e-05,
+      "loss": 1.2783,
+      "step": 26980
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9528449773788452,
+      "learning_rate": 4.287047316944224e-05,
+      "loss": 1.2772,
+      "step": 26981
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.761107325553894,
+      "learning_rate": 4.287020882897172e-05,
+      "loss": 1.2669,
+      "step": 26982
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.954708218574524,
+      "learning_rate": 4.286994448850119e-05,
+      "loss": 1.302,
+      "step": 26983
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.251499891281128,
+      "learning_rate": 4.286968014803067e-05,
+      "loss": 1.3371,
+      "step": 26984
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.575517177581787,
+      "learning_rate": 4.2869415807560135e-05,
+      "loss": 1.4225,
+      "step": 26985
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.7774221897125244,
+      "learning_rate": 4.2869151467089616e-05,
+      "loss": 1.4014,
+      "step": 26986
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7249680757522583,
+      "learning_rate": 4.286888712661909e-05,
+      "loss": 1.172,
+      "step": 26987
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.912317156791687,
+      "learning_rate": 4.286862278614856e-05,
+      "loss": 1.2809,
+      "step": 26988
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.6449971199035645,
+      "learning_rate": 4.2868358445678036e-05,
+      "loss": 1.5591,
+      "step": 26989
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.5175836086273193,
+      "learning_rate": 4.286809410520751e-05,
+      "loss": 1.6302,
+      "step": 26990
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8777049779891968,
+      "learning_rate": 4.2867829764736984e-05,
+      "loss": 1.3155,
+      "step": 26991
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.264342784881592,
+      "learning_rate": 4.286756542426646e-05,
+      "loss": 1.3181,
+      "step": 26992
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6567399501800537,
+      "learning_rate": 4.286730108379593e-05,
+      "loss": 1.266,
+      "step": 26993
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1720387935638428,
+      "learning_rate": 4.2867036743325404e-05,
+      "loss": 1.418,
+      "step": 26994
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0031790733337402,
+      "learning_rate": 4.286677240285488e-05,
+      "loss": 1.2257,
+      "step": 26995
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.915684461593628,
+      "learning_rate": 4.286650806238435e-05,
+      "loss": 1.2779,
+      "step": 26996
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.627168893814087,
+      "learning_rate": 4.286624372191383e-05,
+      "loss": 1.0965,
+      "step": 26997
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.338059425354004,
+      "learning_rate": 4.28659793814433e-05,
+      "loss": 1.3068,
+      "step": 26998
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8184088468551636,
+      "learning_rate": 4.286571504097278e-05,
+      "loss": 1.1635,
+      "step": 26999
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.915626883506775,
+      "learning_rate": 4.2865450700502246e-05,
+      "loss": 1.298,
+      "step": 27000
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8743023872375488,
+      "learning_rate": 4.286518636003172e-05,
+      "loss": 1.3169,
+      "step": 27001
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9794048070907593,
+      "learning_rate": 4.28649220195612e-05,
+      "loss": 1.2556,
+      "step": 27002
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.2477777004241943,
+      "learning_rate": 4.2864657679090666e-05,
+      "loss": 1.3415,
+      "step": 27003
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.942448377609253,
+      "learning_rate": 4.286439333862015e-05,
+      "loss": 1.1828,
+      "step": 27004
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.142261505126953,
+      "learning_rate": 4.2864128998149614e-05,
+      "loss": 1.2935,
+      "step": 27005
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6580232381820679,
+      "learning_rate": 4.2863864657679094e-05,
+      "loss": 1.3428,
+      "step": 27006
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9494656324386597,
+      "learning_rate": 4.286360031720857e-05,
+      "loss": 1.2102,
+      "step": 27007
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1267313957214355,
+      "learning_rate": 4.286333597673804e-05,
+      "loss": 1.2913,
+      "step": 27008
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6561410427093506,
+      "learning_rate": 4.2863071636267515e-05,
+      "loss": 1.1362,
+      "step": 27009
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.5189054012298584,
+      "learning_rate": 4.286280729579699e-05,
+      "loss": 1.3475,
+      "step": 27010
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9738950729370117,
+      "learning_rate": 4.286254295532646e-05,
+      "loss": 1.1922,
+      "step": 27011
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9482438564300537,
+      "learning_rate": 4.2862278614855935e-05,
+      "loss": 1.4357,
+      "step": 27012
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.366259813308716,
+      "learning_rate": 4.286201427438541e-05,
+      "loss": 1.3629,
+      "step": 27013
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3346188068389893,
+      "learning_rate": 4.286174993391488e-05,
+      "loss": 1.3938,
+      "step": 27014
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.9485793113708496,
+      "learning_rate": 4.2861485593444356e-05,
+      "loss": 1.3794,
+      "step": 27015
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.928524136543274,
+      "learning_rate": 4.286122125297383e-05,
+      "loss": 1.3237,
+      "step": 27016
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.66404390335083,
+      "learning_rate": 4.286095691250331e-05,
+      "loss": 1.1258,
+      "step": 27017
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.79163658618927,
+      "learning_rate": 4.286069257203278e-05,
+      "loss": 1.3636,
+      "step": 27018
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.841825008392334,
+      "learning_rate": 4.286042823156226e-05,
+      "loss": 1.3362,
+      "step": 27019
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.011850118637085,
+      "learning_rate": 4.2860163891091724e-05,
+      "loss": 1.9522,
+      "step": 27020
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.0680289268493652,
+      "learning_rate": 4.2859899550621204e-05,
+      "loss": 1.4949,
+      "step": 27021
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9640380144119263,
+      "learning_rate": 4.285963521015068e-05,
+      "loss": 1.3531,
+      "step": 27022
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8480275869369507,
+      "learning_rate": 4.285937086968015e-05,
+      "loss": 1.2384,
+      "step": 27023
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0210154056549072,
+      "learning_rate": 4.2859106529209625e-05,
+      "loss": 1.3306,
+      "step": 27024
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.622499942779541,
+      "learning_rate": 4.28588421887391e-05,
+      "loss": 1.3067,
+      "step": 27025
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.255408763885498,
+      "learning_rate": 4.285857784826857e-05,
+      "loss": 1.2186,
+      "step": 27026
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9589402675628662,
+      "learning_rate": 4.2858313507798046e-05,
+      "loss": 1.3008,
+      "step": 27027
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9508768320083618,
+      "learning_rate": 4.285804916732752e-05,
+      "loss": 1.2429,
+      "step": 27028
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.5804946422576904,
+      "learning_rate": 4.285778482685699e-05,
+      "loss": 1.3881,
+      "step": 27029
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0330562591552734,
+      "learning_rate": 4.2857520486386466e-05,
+      "loss": 1.2956,
+      "step": 27030
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.6072499752044678,
+      "learning_rate": 4.285725614591594e-05,
+      "loss": 1.4864,
+      "step": 27031
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.036630630493164,
+      "learning_rate": 4.285699180544542e-05,
+      "loss": 1.2601,
+      "step": 27032
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7985531091690063,
+      "learning_rate": 4.285672746497489e-05,
+      "loss": 1.266,
+      "step": 27033
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.5817630290985107,
+      "learning_rate": 4.285646312450437e-05,
+      "loss": 1.2191,
+      "step": 27034
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3269240856170654,
+      "learning_rate": 4.2856198784033834e-05,
+      "loss": 1.3182,
+      "step": 27035
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0753023624420166,
+      "learning_rate": 4.2855934443563315e-05,
+      "loss": 1.2324,
+      "step": 27036
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9460171461105347,
+      "learning_rate": 4.285567010309279e-05,
+      "loss": 1.178,
+      "step": 27037
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7061647176742554,
+      "learning_rate": 4.2855405762622255e-05,
+      "loss": 1.2374,
+      "step": 27038
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.656330108642578,
+      "learning_rate": 4.2855141422151735e-05,
+      "loss": 1.3096,
+      "step": 27039
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7815290689468384,
+      "learning_rate": 4.28548770816812e-05,
+      "loss": 1.2481,
+      "step": 27040
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1437485218048096,
+      "learning_rate": 4.285461274121068e-05,
+      "loss": 1.412,
+      "step": 27041
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.997019648551941,
+      "learning_rate": 4.2854348400740156e-05,
+      "loss": 1.2561,
+      "step": 27042
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.4527747631073,
+      "learning_rate": 4.285408406026963e-05,
+      "loss": 1.5396,
+      "step": 27043
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.113729238510132,
+      "learning_rate": 4.28538197197991e-05,
+      "loss": 1.3148,
+      "step": 27044
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1628780364990234,
+      "learning_rate": 4.285355537932858e-05,
+      "loss": 1.3331,
+      "step": 27045
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9537627696990967,
+      "learning_rate": 4.285329103885805e-05,
+      "loss": 1.2179,
+      "step": 27046
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3869333267211914,
+      "learning_rate": 4.2853026698387524e-05,
+      "loss": 1.2384,
+      "step": 27047
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.081109046936035,
+      "learning_rate": 4.2852762357917e-05,
+      "loss": 1.2387,
+      "step": 27048
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0116121768951416,
+      "learning_rate": 4.285249801744647e-05,
+      "loss": 1.3495,
+      "step": 27049
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9243513345718384,
+      "learning_rate": 4.2852233676975945e-05,
+      "loss": 1.2753,
+      "step": 27050
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.960610270500183,
+      "learning_rate": 4.285196933650542e-05,
+      "loss": 1.3079,
+      "step": 27051
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.264845132827759,
+      "learning_rate": 4.28517049960349e-05,
+      "loss": 1.5996,
+      "step": 27052
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.196484327316284,
+      "learning_rate": 4.2851440655564365e-05,
+      "loss": 1.3518,
+      "step": 27053
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.002194881439209,
+      "learning_rate": 4.2851176315093846e-05,
+      "loss": 1.2602,
+      "step": 27054
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9610795974731445,
+      "learning_rate": 4.285091197462331e-05,
+      "loss": 1.2015,
+      "step": 27055
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.612450361251831,
+      "learning_rate": 4.285064763415279e-05,
+      "loss": 1.4041,
+      "step": 27056
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7812883853912354,
+      "learning_rate": 4.2850383293682266e-05,
+      "loss": 1.2656,
+      "step": 27057
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.2533600330352783,
+      "learning_rate": 4.285011895321174e-05,
+      "loss": 1.4713,
+      "step": 27058
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7886141538619995,
+      "learning_rate": 4.2849854612741214e-05,
+      "loss": 1.8835,
+      "step": 27059
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7775272130966187,
+      "learning_rate": 4.284959027227069e-05,
+      "loss": 1.0628,
+      "step": 27060
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9226417541503906,
+      "learning_rate": 4.284932593180016e-05,
+      "loss": 1.2732,
+      "step": 27061
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1256296634674072,
+      "learning_rate": 4.2849061591329634e-05,
+      "loss": 1.3978,
+      "step": 27062
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.362942695617676,
+      "learning_rate": 4.284879725085911e-05,
+      "loss": 1.8262,
+      "step": 27063
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0650970935821533,
+      "learning_rate": 4.284853291038858e-05,
+      "loss": 1.2769,
+      "step": 27064
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.2399537563323975,
+      "learning_rate": 4.2848268569918055e-05,
+      "loss": 1.7226,
+      "step": 27065
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1589057445526123,
+      "learning_rate": 4.284800422944753e-05,
+      "loss": 1.3932,
+      "step": 27066
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9475054740905762,
+      "learning_rate": 4.284773988897701e-05,
+      "loss": 1.3792,
+      "step": 27067
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8037135601043701,
+      "learning_rate": 4.2847475548506476e-05,
+      "loss": 1.0914,
+      "step": 27068
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.434077262878418,
+      "learning_rate": 4.2847211208035956e-05,
+      "loss": 1.2227,
+      "step": 27069
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.5518609285354614,
+      "learning_rate": 4.284694686756542e-05,
+      "loss": 1.2157,
+      "step": 27070
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.734468460083008,
+      "learning_rate": 4.28466825270949e-05,
+      "loss": 1.2598,
+      "step": 27071
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.6120352745056152,
+      "learning_rate": 4.284641818662438e-05,
+      "loss": 1.4582,
+      "step": 27072
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0309338569641113,
+      "learning_rate": 4.284615384615385e-05,
+      "loss": 1.1625,
+      "step": 27073
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0966403484344482,
+      "learning_rate": 4.2845889505683324e-05,
+      "loss": 1.404,
+      "step": 27074
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.835081934928894,
+      "learning_rate": 4.284562516521279e-05,
+      "loss": 1.2869,
+      "step": 27075
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0718400478363037,
+      "learning_rate": 4.284536082474227e-05,
+      "loss": 1.3791,
+      "step": 27076
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9277923107147217,
+      "learning_rate": 4.2845096484271745e-05,
+      "loss": 1.2247,
+      "step": 27077
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.624495267868042,
+      "learning_rate": 4.284483214380122e-05,
+      "loss": 1.5197,
+      "step": 27078
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.991113305091858,
+      "learning_rate": 4.284456780333069e-05,
+      "loss": 1.3943,
+      "step": 27079
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3541066646575928,
+      "learning_rate": 4.2844303462860165e-05,
+      "loss": 1.3419,
+      "step": 27080
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.048062324523926,
+      "learning_rate": 4.284403912238964e-05,
+      "loss": 1.4638,
+      "step": 27081
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.4805495738983154,
+      "learning_rate": 4.284377478191912e-05,
+      "loss": 1.5699,
+      "step": 27082
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1810083389282227,
+      "learning_rate": 4.2843510441448586e-05,
+      "loss": 1.2853,
+      "step": 27083
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7875267267227173,
+      "learning_rate": 4.284324610097806e-05,
+      "loss": 1.1283,
+      "step": 27084
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7852420806884766,
+      "learning_rate": 4.284298176050753e-05,
+      "loss": 1.2904,
+      "step": 27085
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6999503374099731,
+      "learning_rate": 4.284271742003701e-05,
+      "loss": 1.1303,
+      "step": 27086
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8997553586959839,
+      "learning_rate": 4.284245307956649e-05,
+      "loss": 1.331,
+      "step": 27087
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.051649570465088,
+      "learning_rate": 4.2842188739095954e-05,
+      "loss": 1.4108,
+      "step": 27088
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1293210983276367,
+      "learning_rate": 4.2841924398625434e-05,
+      "loss": 1.3592,
+      "step": 27089
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.127504825592041,
+      "learning_rate": 4.28416600581549e-05,
+      "loss": 1.3439,
+      "step": 27090
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.373131513595581,
+      "learning_rate": 4.284139571768438e-05,
+      "loss": 1.2582,
+      "step": 27091
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.757833480834961,
+      "learning_rate": 4.2841131377213855e-05,
+      "loss": 1.1045,
+      "step": 27092
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0428695678710938,
+      "learning_rate": 4.284086703674333e-05,
+      "loss": 1.3669,
+      "step": 27093
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.101452112197876,
+      "learning_rate": 4.28406026962728e-05,
+      "loss": 1.2342,
+      "step": 27094
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3659651279449463,
+      "learning_rate": 4.2840338355802276e-05,
+      "loss": 1.3064,
+      "step": 27095
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.5452756881713867,
+      "learning_rate": 4.284007401533175e-05,
+      "loss": 1.3956,
+      "step": 27096
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.2404773235321045,
+      "learning_rate": 4.283980967486122e-05,
+      "loss": 1.2809,
+      "step": 27097
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7521817684173584,
+      "learning_rate": 4.2839545334390696e-05,
+      "loss": 1.1684,
+      "step": 27098
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.074786424636841,
+      "learning_rate": 4.283928099392017e-05,
+      "loss": 1.4086,
+      "step": 27099
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1662380695343018,
+      "learning_rate": 4.2839016653449644e-05,
+      "loss": 1.3066,
+      "step": 27100
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0791800022125244,
+      "learning_rate": 4.283875231297912e-05,
+      "loss": 1.1899,
+      "step": 27101
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8092303276062012,
+      "learning_rate": 4.28384879725086e-05,
+      "loss": 1.3706,
+      "step": 27102
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.984724760055542,
+      "learning_rate": 4.2838223632038064e-05,
+      "loss": 1.0984,
+      "step": 27103
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.857864260673523,
+      "learning_rate": 4.2837959291567545e-05,
+      "loss": 1.327,
+      "step": 27104
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6154214143753052,
+      "learning_rate": 4.283769495109701e-05,
+      "loss": 1.2788,
+      "step": 27105
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.119006395339966,
+      "learning_rate": 4.283743061062649e-05,
+      "loss": 1.3333,
+      "step": 27106
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8094983100891113,
+      "learning_rate": 4.2837166270155965e-05,
+      "loss": 1.1344,
+      "step": 27107
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.098926067352295,
+      "learning_rate": 4.283690192968544e-05,
+      "loss": 1.3479,
+      "step": 27108
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9415090084075928,
+      "learning_rate": 4.283663758921491e-05,
+      "loss": 1.3009,
+      "step": 27109
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.5761550664901733,
+      "learning_rate": 4.283637324874438e-05,
+      "loss": 1.7049,
+      "step": 27110
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3210012912750244,
+      "learning_rate": 4.283610890827386e-05,
+      "loss": 1.4153,
+      "step": 27111
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.7573204040527344,
+      "learning_rate": 4.283584456780333e-05,
+      "loss": 1.152,
+      "step": 27112
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7182451486587524,
+      "learning_rate": 4.283558022733281e-05,
+      "loss": 1.1096,
+      "step": 27113
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.2494399547576904,
+      "learning_rate": 4.283531588686228e-05,
+      "loss": 1.3716,
+      "step": 27114
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9929014444351196,
+      "learning_rate": 4.2835051546391754e-05,
+      "loss": 1.2986,
+      "step": 27115
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.6319313049316406,
+      "learning_rate": 4.283478720592123e-05,
+      "loss": 1.2663,
+      "step": 27116
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.744332194328308,
+      "learning_rate": 4.283452286545071e-05,
+      "loss": 1.1913,
+      "step": 27117
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1063926219940186,
+      "learning_rate": 4.2834258524980175e-05,
+      "loss": 1.1547,
+      "step": 27118
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.2197084426879883,
+      "learning_rate": 4.283399418450965e-05,
+      "loss": 1.2426,
+      "step": 27119
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.06569766998291,
+      "learning_rate": 4.283372984403912e-05,
+      "loss": 1.2284,
+      "step": 27120
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9752886295318604,
+      "learning_rate": 4.2833465503568595e-05,
+      "loss": 1.187,
+      "step": 27121
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.6285576820373535,
+      "learning_rate": 4.2833201163098076e-05,
+      "loss": 1.4811,
+      "step": 27122
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.141331911087036,
+      "learning_rate": 4.283293682262754e-05,
+      "loss": 1.3473,
+      "step": 27123
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.2060885429382324,
+      "learning_rate": 4.283267248215702e-05,
+      "loss": 1.4169,
+      "step": 27124
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.4808223247528076,
+      "learning_rate": 4.283240814168649e-05,
+      "loss": 1.4154,
+      "step": 27125
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.838012456893921,
+      "learning_rate": 4.283214380121597e-05,
+      "loss": 1.3229,
+      "step": 27126
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9728691577911377,
+      "learning_rate": 4.2831879460745443e-05,
+      "loss": 1.2536,
+      "step": 27127
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7145687341690063,
+      "learning_rate": 4.283161512027492e-05,
+      "loss": 1.1966,
+      "step": 27128
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.170346260070801,
+      "learning_rate": 4.283135077980439e-05,
+      "loss": 1.209,
+      "step": 27129
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3292250633239746,
+      "learning_rate": 4.2831086439333864e-05,
+      "loss": 1.2599,
+      "step": 27130
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.2056469917297363,
+      "learning_rate": 4.283082209886334e-05,
+      "loss": 1.2913,
+      "step": 27131
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.016946315765381,
+      "learning_rate": 4.283055775839281e-05,
+      "loss": 1.2345,
+      "step": 27132
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9226819276809692,
+      "learning_rate": 4.2830293417922285e-05,
+      "loss": 1.2874,
+      "step": 27133
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.414219617843628,
+      "learning_rate": 4.283002907745176e-05,
+      "loss": 1.2685,
+      "step": 27134
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6937711238861084,
+      "learning_rate": 4.282976473698123e-05,
+      "loss": 1.1931,
+      "step": 27135
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.178504228591919,
+      "learning_rate": 4.2829500396510706e-05,
+      "loss": 1.3029,
+      "step": 27136
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.457305908203125,
+      "learning_rate": 4.2829236056040186e-05,
+      "loss": 1.1349,
+      "step": 27137
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9782423973083496,
+      "learning_rate": 4.282897171556965e-05,
+      "loss": 1.3362,
+      "step": 27138
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0265462398529053,
+      "learning_rate": 4.282870737509913e-05,
+      "loss": 1.3893,
+      "step": 27139
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0785529613494873,
+      "learning_rate": 4.28284430346286e-05,
+      "loss": 1.365,
+      "step": 27140
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.5265207290649414,
+      "learning_rate": 4.282817869415808e-05,
+      "loss": 1.5457,
+      "step": 27141
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.35394549369812,
+      "learning_rate": 4.2827914353687554e-05,
+      "loss": 1.4305,
+      "step": 27142
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.4822733402252197,
+      "learning_rate": 4.282765001321703e-05,
+      "loss": 1.2257,
+      "step": 27143
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7031574249267578,
+      "learning_rate": 4.28273856727465e-05,
+      "loss": 1.2596,
+      "step": 27144
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.389655113220215,
+      "learning_rate": 4.2827121332275975e-05,
+      "loss": 1.3715,
+      "step": 27145
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.136338233947754,
+      "learning_rate": 4.282685699180545e-05,
+      "loss": 1.3043,
+      "step": 27146
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.4700441360473633,
+      "learning_rate": 4.282659265133492e-05,
+      "loss": 1.3958,
+      "step": 27147
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.751184344291687,
+      "learning_rate": 4.2826328310864395e-05,
+      "loss": 1.2972,
+      "step": 27148
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.597073554992676,
+      "learning_rate": 4.282606397039387e-05,
+      "loss": 1.5239,
+      "step": 27149
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3960814476013184,
+      "learning_rate": 4.282579962992334e-05,
+      "loss": 1.3899,
+      "step": 27150
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9886174201965332,
+      "learning_rate": 4.2825535289452816e-05,
+      "loss": 1.2715,
+      "step": 27151
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.763287305831909,
+      "learning_rate": 4.2825270948982296e-05,
+      "loss": 1.4419,
+      "step": 27152
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3396384716033936,
+      "learning_rate": 4.282500660851176e-05,
+      "loss": 1.4999,
+      "step": 27153
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9878495931625366,
+      "learning_rate": 4.2824742268041243e-05,
+      "loss": 1.2881,
+      "step": 27154
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9657779932022095,
+      "learning_rate": 4.282447792757071e-05,
+      "loss": 1.3963,
+      "step": 27155
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.375321388244629,
+      "learning_rate": 4.2824213587100184e-05,
+      "loss": 1.3611,
+      "step": 27156
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7468432188034058,
+      "learning_rate": 4.2823949246629664e-05,
+      "loss": 1.2494,
+      "step": 27157
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9031494855880737,
+      "learning_rate": 4.282368490615913e-05,
+      "loss": 1.2057,
+      "step": 27158
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8822863101959229,
+      "learning_rate": 4.282342056568861e-05,
+      "loss": 1.2807,
+      "step": 27159
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9795093536376953,
+      "learning_rate": 4.282315622521808e-05,
+      "loss": 1.3269,
+      "step": 27160
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.5850937366485596,
+      "learning_rate": 4.282289188474756e-05,
+      "loss": 1.3927,
+      "step": 27161
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6312166452407837,
+      "learning_rate": 4.282262754427703e-05,
+      "loss": 1.1835,
+      "step": 27162
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.049736499786377,
+      "learning_rate": 4.2822363203806506e-05,
+      "loss": 1.4014,
+      "step": 27163
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6702038049697876,
+      "learning_rate": 4.282209886333598e-05,
+      "loss": 2.2733,
+      "step": 27164
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2827110290527344,
+      "learning_rate": 4.282183452286545e-05,
+      "loss": 1.4085,
+      "step": 27165
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3179962635040283,
+      "learning_rate": 4.2821570182394926e-05,
+      "loss": 1.3955,
+      "step": 27166
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.67162823677063,
+      "learning_rate": 4.28213058419244e-05,
+      "loss": 1.3575,
+      "step": 27167
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0499682426452637,
+      "learning_rate": 4.2821041501453873e-05,
+      "loss": 1.3894,
+      "step": 27168
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9755231142044067,
+      "learning_rate": 4.282077716098335e-05,
+      "loss": 1.3788,
+      "step": 27169
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3776347637176514,
+      "learning_rate": 4.282051282051282e-05,
+      "loss": 1.3339,
+      "step": 27170
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.678540825843811,
+      "learning_rate": 4.2820248480042294e-05,
+      "loss": 1.1575,
+      "step": 27171
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9096314907073975,
+      "learning_rate": 4.2819984139571775e-05,
+      "loss": 1.3385,
+      "step": 27172
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6806285381317139,
+      "learning_rate": 4.281971979910124e-05,
+      "loss": 1.2258,
+      "step": 27173
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0350165367126465,
+      "learning_rate": 4.281945545863072e-05,
+      "loss": 1.3304,
+      "step": 27174
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.315431833267212,
+      "learning_rate": 4.281919111816019e-05,
+      "loss": 1.4074,
+      "step": 27175
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8905327320098877,
+      "learning_rate": 4.281892677768967e-05,
+      "loss": 1.2104,
+      "step": 27176
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.0786349773406982,
+      "learning_rate": 4.281866243721914e-05,
+      "loss": 1.4256,
+      "step": 27177
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6213841438293457,
+      "learning_rate": 4.2818398096748616e-05,
+      "loss": 1.233,
+      "step": 27178
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.023693323135376,
+      "learning_rate": 4.281813375627809e-05,
+      "loss": 1.1176,
+      "step": 27179
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9972147941589355,
+      "learning_rate": 4.281786941580756e-05,
+      "loss": 1.287,
+      "step": 27180
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.111607074737549,
+      "learning_rate": 4.281760507533704e-05,
+      "loss": 1.3822,
+      "step": 27181
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0193984508514404,
+      "learning_rate": 4.281734073486651e-05,
+      "loss": 1.3583,
+      "step": 27182
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.332993507385254,
+      "learning_rate": 4.2817076394395984e-05,
+      "loss": 1.4617,
+      "step": 27183
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9526599645614624,
+      "learning_rate": 4.281681205392546e-05,
+      "loss": 1.4027,
+      "step": 27184
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.14567494392395,
+      "learning_rate": 4.281654771345493e-05,
+      "loss": 1.2806,
+      "step": 27185
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2135133743286133,
+      "learning_rate": 4.2816283372984404e-05,
+      "loss": 1.4746,
+      "step": 27186
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.988952398300171,
+      "learning_rate": 4.2816019032513885e-05,
+      "loss": 1.6022,
+      "step": 27187
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.1911725997924805,
+      "learning_rate": 4.281575469204335e-05,
+      "loss": 1.4296,
+      "step": 27188
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8120988607406616,
+      "learning_rate": 4.281549035157283e-05,
+      "loss": 1.27,
+      "step": 27189
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.412071466445923,
+      "learning_rate": 4.28152260111023e-05,
+      "loss": 1.1853,
+      "step": 27190
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1844465732574463,
+      "learning_rate": 4.281496167063178e-05,
+      "loss": 1.1129,
+      "step": 27191
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9566912651062012,
+      "learning_rate": 4.281469733016125e-05,
+      "loss": 1.3694,
+      "step": 27192
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.205535411834717,
+      "learning_rate": 4.281443298969072e-05,
+      "loss": 1.3745,
+      "step": 27193
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.087113618850708,
+      "learning_rate": 4.28141686492202e-05,
+      "loss": 1.5191,
+      "step": 27194
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7359498739242554,
+      "learning_rate": 4.281390430874967e-05,
+      "loss": 1.1376,
+      "step": 27195
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.377575397491455,
+      "learning_rate": 4.281363996827915e-05,
+      "loss": 1.4031,
+      "step": 27196
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.172665596008301,
+      "learning_rate": 4.281337562780862e-05,
+      "loss": 1.3471,
+      "step": 27197
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2307369709014893,
+      "learning_rate": 4.2813111287338094e-05,
+      "loss": 1.6115,
+      "step": 27198
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1479663848876953,
+      "learning_rate": 4.281284694686757e-05,
+      "loss": 1.4501,
+      "step": 27199
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.209946870803833,
+      "learning_rate": 4.281258260639704e-05,
+      "loss": 1.3318,
+      "step": 27200
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1770341396331787,
+      "learning_rate": 4.2812318265926515e-05,
+      "loss": 1.3089,
+      "step": 27201
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7074353694915771,
+      "learning_rate": 4.281205392545599e-05,
+      "loss": 1.2688,
+      "step": 27202
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2184150218963623,
+      "learning_rate": 4.281178958498546e-05,
+      "loss": 1.391,
+      "step": 27203
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8584691286087036,
+      "learning_rate": 4.2811525244514936e-05,
+      "loss": 1.2959,
+      "step": 27204
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3248000144958496,
+      "learning_rate": 4.281126090404441e-05,
+      "loss": 1.4794,
+      "step": 27205
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.5481722354888916,
+      "learning_rate": 4.281099656357388e-05,
+      "loss": 1.2146,
+      "step": 27206
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7722793817520142,
+      "learning_rate": 4.281073222310336e-05,
+      "loss": 1.2224,
+      "step": 27207
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.78787612915039,
+      "learning_rate": 4.281046788263283e-05,
+      "loss": 1.7457,
+      "step": 27208
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1827569007873535,
+      "learning_rate": 4.281020354216231e-05,
+      "loss": 1.3944,
+      "step": 27209
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.109692096710205,
+      "learning_rate": 4.280993920169178e-05,
+      "loss": 1.3482,
+      "step": 27210
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8666434288024902,
+      "learning_rate": 4.280967486122126e-05,
+      "loss": 1.25,
+      "step": 27211
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.4608752727508545,
+      "learning_rate": 4.280941052075073e-05,
+      "loss": 1.4254,
+      "step": 27212
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8352019786834717,
+      "learning_rate": 4.2809146180280204e-05,
+      "loss": 1.2417,
+      "step": 27213
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9196933507919312,
+      "learning_rate": 4.280888183980968e-05,
+      "loss": 1.1167,
+      "step": 27214
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.139477491378784,
+      "learning_rate": 4.280861749933915e-05,
+      "loss": 1.3319,
+      "step": 27215
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0869557857513428,
+      "learning_rate": 4.2808353158868625e-05,
+      "loss": 1.1493,
+      "step": 27216
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9966672658920288,
+      "learning_rate": 4.28080888183981e-05,
+      "loss": 1.3589,
+      "step": 27217
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.747351169586182,
+      "learning_rate": 4.280782447792757e-05,
+      "loss": 1.4694,
+      "step": 27218
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.135408401489258,
+      "learning_rate": 4.2807560137457046e-05,
+      "loss": 1.2623,
+      "step": 27219
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8487412929534912,
+      "learning_rate": 4.280729579698652e-05,
+      "loss": 1.3268,
+      "step": 27220
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6984550952911377,
+      "learning_rate": 4.280703145651599e-05,
+      "loss": 1.3073,
+      "step": 27221
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.192221164703369,
+      "learning_rate": 4.280676711604547e-05,
+      "loss": 1.429,
+      "step": 27222
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6242380142211914,
+      "learning_rate": 4.280650277557494e-05,
+      "loss": 1.638,
+      "step": 27223
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9183655977249146,
+      "learning_rate": 4.280623843510442e-05,
+      "loss": 1.4723,
+      "step": 27224
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1494784355163574,
+      "learning_rate": 4.280597409463389e-05,
+      "loss": 1.3389,
+      "step": 27225
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.311634063720703,
+      "learning_rate": 4.280570975416337e-05,
+      "loss": 1.4717,
+      "step": 27226
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.6067519187927246,
+      "learning_rate": 4.280544541369284e-05,
+      "loss": 1.3555,
+      "step": 27227
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7570607662200928,
+      "learning_rate": 4.280518107322231e-05,
+      "loss": 1.2494,
+      "step": 27228
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9770734310150146,
+      "learning_rate": 4.280491673275179e-05,
+      "loss": 1.3229,
+      "step": 27229
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2434706687927246,
+      "learning_rate": 4.2804652392281255e-05,
+      "loss": 1.246,
+      "step": 27230
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0152313709259033,
+      "learning_rate": 4.2804388051810736e-05,
+      "loss": 1.301,
+      "step": 27231
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.736743450164795,
+      "learning_rate": 4.280412371134021e-05,
+      "loss": 1.4294,
+      "step": 27232
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.219289302825928,
+      "learning_rate": 4.280385937086968e-05,
+      "loss": 1.717,
+      "step": 27233
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6996350288391113,
+      "learning_rate": 4.2803595030399156e-05,
+      "loss": 1.2492,
+      "step": 27234
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.264164447784424,
+      "learning_rate": 4.280333068992863e-05,
+      "loss": 1.3562,
+      "step": 27235
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2824909687042236,
+      "learning_rate": 4.28030663494581e-05,
+      "loss": 1.2797,
+      "step": 27236
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.991188406944275,
+      "learning_rate": 4.280280200898758e-05,
+      "loss": 1.244,
+      "step": 27237
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1055586338043213,
+      "learning_rate": 4.280253766851705e-05,
+      "loss": 1.2913,
+      "step": 27238
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0976321697235107,
+      "learning_rate": 4.2802273328046524e-05,
+      "loss": 1.3385,
+      "step": 27239
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.696582555770874,
+      "learning_rate": 4.2802008987576e-05,
+      "loss": 1.266,
+      "step": 27240
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.950943946838379,
+      "learning_rate": 4.280174464710547e-05,
+      "loss": 1.3096,
+      "step": 27241
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.018345832824707,
+      "learning_rate": 4.280148030663495e-05,
+      "loss": 1.193,
+      "step": 27242
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.839332342147827,
+      "learning_rate": 4.280121596616442e-05,
+      "loss": 1.5751,
+      "step": 27243
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.005791664123535,
+      "learning_rate": 4.28009516256939e-05,
+      "loss": 1.278,
+      "step": 27244
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1372146606445312,
+      "learning_rate": 4.2800687285223366e-05,
+      "loss": 1.4489,
+      "step": 27245
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.23494815826416,
+      "learning_rate": 4.2800422944752846e-05,
+      "loss": 1.3009,
+      "step": 27246
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.6701126098632812,
+      "learning_rate": 4.280015860428232e-05,
+      "loss": 1.6439,
+      "step": 27247
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8021281957626343,
+      "learning_rate": 4.279989426381179e-05,
+      "loss": 1.137,
+      "step": 27248
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9183156490325928,
+      "learning_rate": 4.2799629923341267e-05,
+      "loss": 1.3451,
+      "step": 27249
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.279714345932007,
+      "learning_rate": 4.279936558287074e-05,
+      "loss": 1.2683,
+      "step": 27250
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.969686508178711,
+      "learning_rate": 4.2799101242400214e-05,
+      "loss": 1.4722,
+      "step": 27251
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9975879192352295,
+      "learning_rate": 4.279883690192969e-05,
+      "loss": 1.2934,
+      "step": 27252
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.6335320472717285,
+      "learning_rate": 4.279857256145916e-05,
+      "loss": 1.472,
+      "step": 27253
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1573715209960938,
+      "learning_rate": 4.2798308220988634e-05,
+      "loss": 1.2829,
+      "step": 27254
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.917915940284729,
+      "learning_rate": 4.279804388051811e-05,
+      "loss": 1.3399,
+      "step": 27255
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.073214530944824,
+      "learning_rate": 4.279777954004758e-05,
+      "loss": 1.4263,
+      "step": 27256
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0215954780578613,
+      "learning_rate": 4.279751519957706e-05,
+      "loss": 1.2467,
+      "step": 27257
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.866239309310913,
+      "learning_rate": 4.279725085910653e-05,
+      "loss": 1.4384,
+      "step": 27258
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0999906063079834,
+      "learning_rate": 4.279698651863601e-05,
+      "loss": 1.2541,
+      "step": 27259
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.645004153251648,
+      "learning_rate": 4.2796722178165476e-05,
+      "loss": 1.1972,
+      "step": 27260
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.435521125793457,
+      "learning_rate": 4.2796457837694956e-05,
+      "loss": 1.2377,
+      "step": 27261
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0487334728240967,
+      "learning_rate": 4.279619349722443e-05,
+      "loss": 1.3175,
+      "step": 27262
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7445957660675049,
+      "learning_rate": 4.27959291567539e-05,
+      "loss": 1.3275,
+      "step": 27263
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9430373907089233,
+      "learning_rate": 4.279566481628338e-05,
+      "loss": 1.411,
+      "step": 27264
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1844890117645264,
+      "learning_rate": 4.2795400475812844e-05,
+      "loss": 1.2941,
+      "step": 27265
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.132453203201294,
+      "learning_rate": 4.2795136135342324e-05,
+      "loss": 1.3166,
+      "step": 27266
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9772824048995972,
+      "learning_rate": 4.27948717948718e-05,
+      "loss": 1.2079,
+      "step": 27267
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.659524917602539,
+      "learning_rate": 4.279460745440127e-05,
+      "loss": 1.2858,
+      "step": 27268
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.255845785140991,
+      "learning_rate": 4.2794343113930745e-05,
+      "loss": 1.3893,
+      "step": 27269
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1839590072631836,
+      "learning_rate": 4.279407877346022e-05,
+      "loss": 1.3052,
+      "step": 27270
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0993313789367676,
+      "learning_rate": 4.279381443298969e-05,
+      "loss": 1.271,
+      "step": 27271
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8211418390274048,
+      "learning_rate": 4.279355009251917e-05,
+      "loss": 1.0727,
+      "step": 27272
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3353140354156494,
+      "learning_rate": 4.279328575204864e-05,
+      "loss": 1.4602,
+      "step": 27273
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.142175674438477,
+      "learning_rate": 4.279302141157811e-05,
+      "loss": 1.4211,
+      "step": 27274
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0670363903045654,
+      "learning_rate": 4.2792757071107586e-05,
+      "loss": 1.3145,
+      "step": 27275
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8799457550048828,
+      "learning_rate": 4.279249273063706e-05,
+      "loss": 1.2006,
+      "step": 27276
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.080066204071045,
+      "learning_rate": 4.279222839016654e-05,
+      "loss": 1.2873,
+      "step": 27277
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3450541496276855,
+      "learning_rate": 4.279196404969601e-05,
+      "loss": 1.3682,
+      "step": 27278
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.803288221359253,
+      "learning_rate": 4.279169970922549e-05,
+      "loss": 1.2405,
+      "step": 27279
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8480607271194458,
+      "learning_rate": 4.2791435368754954e-05,
+      "loss": 1.3112,
+      "step": 27280
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1984901428222656,
+      "learning_rate": 4.2791171028284434e-05,
+      "loss": 1.2384,
+      "step": 27281
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9239978790283203,
+      "learning_rate": 4.279090668781391e-05,
+      "loss": 2.4405,
+      "step": 27282
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.910929799079895,
+      "learning_rate": 4.279064234734338e-05,
+      "loss": 1.2548,
+      "step": 27283
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9676963090896606,
+      "learning_rate": 4.2790378006872855e-05,
+      "loss": 1.2622,
+      "step": 27284
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1706390380859375,
+      "learning_rate": 4.279011366640233e-05,
+      "loss": 1.3188,
+      "step": 27285
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7668184041976929,
+      "learning_rate": 4.27898493259318e-05,
+      "loss": 1.19,
+      "step": 27286
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9982225894927979,
+      "learning_rate": 4.2789584985461276e-05,
+      "loss": 1.3384,
+      "step": 27287
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9244250059127808,
+      "learning_rate": 4.278932064499075e-05,
+      "loss": 1.1872,
+      "step": 27288
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3559811115264893,
+      "learning_rate": 4.278905630452022e-05,
+      "loss": 1.3954,
+      "step": 27289
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.145228147506714,
+      "learning_rate": 4.2788791964049697e-05,
+      "loss": 1.2098,
+      "step": 27290
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1754472255706787,
+      "learning_rate": 4.278852762357917e-05,
+      "loss": 1.4805,
+      "step": 27291
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.4250903129577637,
+      "learning_rate": 4.278826328310865e-05,
+      "loss": 1.3794,
+      "step": 27292
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8659650087356567,
+      "learning_rate": 4.278799894263812e-05,
+      "loss": 1.2402,
+      "step": 27293
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9927287101745605,
+      "learning_rate": 4.27877346021676e-05,
+      "loss": 1.2795,
+      "step": 27294
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.581747055053711,
+      "learning_rate": 4.2787470261697064e-05,
+      "loss": 1.4321,
+      "step": 27295
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7902209758758545,
+      "learning_rate": 4.2787205921226545e-05,
+      "loss": 1.1402,
+      "step": 27296
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.808358907699585,
+      "learning_rate": 4.278694158075602e-05,
+      "loss": 1.1987,
+      "step": 27297
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0104520320892334,
+      "learning_rate": 4.278667724028549e-05,
+      "loss": 1.2737,
+      "step": 27298
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.9798948764801025,
+      "learning_rate": 4.2786412899814965e-05,
+      "loss": 1.5826,
+      "step": 27299
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9700798988342285,
+      "learning_rate": 4.278614855934443e-05,
+      "loss": 1.5987,
+      "step": 27300
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6028436422348022,
+      "learning_rate": 4.278588421887391e-05,
+      "loss": 1.0546,
+      "step": 27301
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.5911959409713745,
+      "learning_rate": 4.2785619878403386e-05,
+      "loss": 1.2252,
+      "step": 27302
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.956015706062317,
+      "learning_rate": 4.278535553793286e-05,
+      "loss": 1.1978,
+      "step": 27303
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8094596862792969,
+      "learning_rate": 4.278509119746233e-05,
+      "loss": 1.133,
+      "step": 27304
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3876700401306152,
+      "learning_rate": 4.278482685699181e-05,
+      "loss": 1.4538,
+      "step": 27305
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1934962272644043,
+      "learning_rate": 4.278456251652128e-05,
+      "loss": 1.356,
+      "step": 27306
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.019042730331421,
+      "learning_rate": 4.278429817605076e-05,
+      "loss": 1.2604,
+      "step": 27307
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.021329402923584,
+      "learning_rate": 4.278403383558023e-05,
+      "loss": 1.3694,
+      "step": 27308
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8497374057769775,
+      "learning_rate": 4.278376949510971e-05,
+      "loss": 1.2562,
+      "step": 27309
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9045052528381348,
+      "learning_rate": 4.2783505154639175e-05,
+      "loss": 1.3782,
+      "step": 27310
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.369164228439331,
+      "learning_rate": 4.278324081416865e-05,
+      "loss": 1.3721,
+      "step": 27311
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7889535427093506,
+      "learning_rate": 4.278297647369813e-05,
+      "loss": 1.2729,
+      "step": 27312
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.691734552383423,
+      "learning_rate": 4.2782712133227595e-05,
+      "loss": 2.6395,
+      "step": 27313
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7798807621002197,
+      "learning_rate": 4.2782447792757076e-05,
+      "loss": 1.1576,
+      "step": 27314
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6634896993637085,
+      "learning_rate": 4.278218345228654e-05,
+      "loss": 1.9191,
+      "step": 27315
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8686820268630981,
+      "learning_rate": 4.278191911181602e-05,
+      "loss": 1.2341,
+      "step": 27316
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8644670248031616,
+      "learning_rate": 4.2781654771345496e-05,
+      "loss": 1.2621,
+      "step": 27317
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.259284019470215,
+      "learning_rate": 4.278139043087497e-05,
+      "loss": 1.2396,
+      "step": 27318
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.5117499828338623,
+      "learning_rate": 4.2781126090404444e-05,
+      "loss": 1.4285,
+      "step": 27319
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3123180866241455,
+      "learning_rate": 4.278086174993392e-05,
+      "loss": 1.3817,
+      "step": 27320
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.067578077316284,
+      "learning_rate": 4.278059740946339e-05,
+      "loss": 1.1554,
+      "step": 27321
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.563165545463562,
+      "learning_rate": 4.2780333068992864e-05,
+      "loss": 1.1871,
+      "step": 27322
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.5670267343521118,
+      "learning_rate": 4.278006872852234e-05,
+      "loss": 1.1283,
+      "step": 27323
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6390787363052368,
+      "learning_rate": 4.277980438805181e-05,
+      "loss": 1.1072,
+      "step": 27324
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0595219135284424,
+      "learning_rate": 4.2779540047581285e-05,
+      "loss": 1.4093,
+      "step": 27325
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9251267910003662,
+      "learning_rate": 4.277927570711076e-05,
+      "loss": 1.2096,
+      "step": 27326
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.733020544052124,
+      "learning_rate": 4.277901136664024e-05,
+      "loss": 2.2202,
+      "step": 27327
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0851922035217285,
+      "learning_rate": 4.2778747026169706e-05,
+      "loss": 1.2499,
+      "step": 27328
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2513651847839355,
+      "learning_rate": 4.2778482685699186e-05,
+      "loss": 1.3845,
+      "step": 27329
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0009799003601074,
+      "learning_rate": 4.277821834522865e-05,
+      "loss": 1.4094,
+      "step": 27330
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8573708534240723,
+      "learning_rate": 4.277795400475813e-05,
+      "loss": 1.2851,
+      "step": 27331
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9414472579956055,
+      "learning_rate": 4.277768966428761e-05,
+      "loss": 1.3179,
+      "step": 27332
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0743844509124756,
+      "learning_rate": 4.277742532381708e-05,
+      "loss": 1.319,
+      "step": 27333
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.507910966873169,
+      "learning_rate": 4.2777160983346554e-05,
+      "loss": 1.2985,
+      "step": 27334
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3522298336029053,
+      "learning_rate": 4.277689664287603e-05,
+      "loss": 1.4847,
+      "step": 27335
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.6547605991363525,
+      "learning_rate": 4.27766323024055e-05,
+      "loss": 1.453,
+      "step": 27336
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7209566831588745,
+      "learning_rate": 4.2776367961934975e-05,
+      "loss": 1.2176,
+      "step": 27337
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0415420532226562,
+      "learning_rate": 4.277610362146445e-05,
+      "loss": 1.2445,
+      "step": 27338
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9502215385437012,
+      "learning_rate": 4.277583928099392e-05,
+      "loss": 1.2587,
+      "step": 27339
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.697535753250122,
+      "learning_rate": 4.2775574940523395e-05,
+      "loss": 1.3862,
+      "step": 27340
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.7860805988311768,
+      "learning_rate": 4.277531060005287e-05,
+      "loss": 1.4301,
+      "step": 27341
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.138329029083252,
+      "learning_rate": 4.277504625958235e-05,
+      "loss": 1.3753,
+      "step": 27342
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9824870824813843,
+      "learning_rate": 4.2774781919111816e-05,
+      "loss": 1.2406,
+      "step": 27343
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7652307748794556,
+      "learning_rate": 4.2774517578641296e-05,
+      "loss": 1.2512,
+      "step": 27344
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.296297311782837,
+      "learning_rate": 4.277425323817076e-05,
+      "loss": 1.4105,
+      "step": 27345
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.1035614013671875,
+      "learning_rate": 4.277398889770024e-05,
+      "loss": 1.2905,
+      "step": 27346
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9652687311172485,
+      "learning_rate": 4.277372455722972e-05,
+      "loss": 1.3773,
+      "step": 27347
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3866288661956787,
+      "learning_rate": 4.2773460216759184e-05,
+      "loss": 1.4465,
+      "step": 27348
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.5212721824645996,
+      "learning_rate": 4.2773195876288664e-05,
+      "loss": 1.2613,
+      "step": 27349
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4483370780944824,
+      "learning_rate": 4.277293153581813e-05,
+      "loss": 1.3323,
+      "step": 27350
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3168890476226807,
+      "learning_rate": 4.277266719534761e-05,
+      "loss": 1.4049,
+      "step": 27351
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0189220905303955,
+      "learning_rate": 4.2772402854877085e-05,
+      "loss": 1.2983,
+      "step": 27352
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7779022455215454,
+      "learning_rate": 4.277213851440656e-05,
+      "loss": 1.1672,
+      "step": 27353
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.7283639907836914,
+      "learning_rate": 4.277187417393603e-05,
+      "loss": 1.4331,
+      "step": 27354
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.1711905002593994,
+      "learning_rate": 4.2771609833465506e-05,
+      "loss": 1.2692,
+      "step": 27355
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.18571138381958,
+      "learning_rate": 4.277134549299498e-05,
+      "loss": 1.3003,
+      "step": 27356
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9415537118911743,
+      "learning_rate": 4.277108115252445e-05,
+      "loss": 1.3649,
+      "step": 27357
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.919073224067688,
+      "learning_rate": 4.2770816812053926e-05,
+      "loss": 1.3322,
+      "step": 27358
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8426843881607056,
+      "learning_rate": 4.27705524715834e-05,
+      "loss": 1.2736,
+      "step": 27359
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4616432189941406,
+      "learning_rate": 4.2770288131112874e-05,
+      "loss": 1.3106,
+      "step": 27360
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.035161018371582,
+      "learning_rate": 4.277002379064235e-05,
+      "loss": 1.4185,
+      "step": 27361
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.1091957092285156,
+      "learning_rate": 4.276975945017183e-05,
+      "loss": 1.2834,
+      "step": 27362
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0634846687316895,
+      "learning_rate": 4.2769495109701294e-05,
+      "loss": 1.1958,
+      "step": 27363
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4246397018432617,
+      "learning_rate": 4.2769230769230775e-05,
+      "loss": 1.2915,
+      "step": 27364
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.1827831268310547,
+      "learning_rate": 4.276896642876024e-05,
+      "loss": 1.3056,
+      "step": 27365
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8387503623962402,
+      "learning_rate": 4.276870208828972e-05,
+      "loss": 1.2774,
+      "step": 27366
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.237882137298584,
+      "learning_rate": 4.2768437747819195e-05,
+      "loss": 1.4426,
+      "step": 27367
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0962765216827393,
+      "learning_rate": 4.276817340734867e-05,
+      "loss": 1.2962,
+      "step": 27368
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.777531862258911,
+      "learning_rate": 4.276790906687814e-05,
+      "loss": 1.4327,
+      "step": 27369
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.973894476890564,
+      "learning_rate": 4.2767644726407616e-05,
+      "loss": 1.2839,
+      "step": 27370
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6400424242019653,
+      "learning_rate": 4.276738038593709e-05,
+      "loss": 1.2488,
+      "step": 27371
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.5740437507629395,
+      "learning_rate": 4.276711604546656e-05,
+      "loss": 1.5489,
+      "step": 27372
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.9691407680511475,
+      "learning_rate": 4.276685170499604e-05,
+      "loss": 1.4483,
+      "step": 27373
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.542337417602539,
+      "learning_rate": 4.276658736452551e-05,
+      "loss": 1.2398,
+      "step": 27374
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.184951066970825,
+      "learning_rate": 4.2766323024054984e-05,
+      "loss": 1.3661,
+      "step": 27375
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0087060928344727,
+      "learning_rate": 4.276605868358446e-05,
+      "loss": 1.3749,
+      "step": 27376
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0297694206237793,
+      "learning_rate": 4.276579434311394e-05,
+      "loss": 1.2939,
+      "step": 27377
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.025022029876709,
+      "learning_rate": 4.2765530002643405e-05,
+      "loss": 1.7258,
+      "step": 27378
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.237178325653076,
+      "learning_rate": 4.2765265662172885e-05,
+      "loss": 1.3178,
+      "step": 27379
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.055867910385132,
+      "learning_rate": 4.276500132170235e-05,
+      "loss": 1.3233,
+      "step": 27380
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7571864128112793,
+      "learning_rate": 4.276473698123183e-05,
+      "loss": 1.1549,
+      "step": 27381
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.46337890625,
+      "learning_rate": 4.2764472640761306e-05,
+      "loss": 1.3017,
+      "step": 27382
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.315690517425537,
+      "learning_rate": 4.276420830029077e-05,
+      "loss": 1.3634,
+      "step": 27383
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9955617189407349,
+      "learning_rate": 4.276394395982025e-05,
+      "loss": 1.2055,
+      "step": 27384
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7434555292129517,
+      "learning_rate": 4.276367961934972e-05,
+      "loss": 1.2014,
+      "step": 27385
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0882349014282227,
+      "learning_rate": 4.27634152788792e-05,
+      "loss": 1.2641,
+      "step": 27386
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.7588629722595215,
+      "learning_rate": 4.2763150938408674e-05,
+      "loss": 1.3928,
+      "step": 27387
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.880152702331543,
+      "learning_rate": 4.276288659793815e-05,
+      "loss": 1.299,
+      "step": 27388
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9901516437530518,
+      "learning_rate": 4.276262225746762e-05,
+      "loss": 1.1095,
+      "step": 27389
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.135636329650879,
+      "learning_rate": 4.2762357916997094e-05,
+      "loss": 1.3666,
+      "step": 27390
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.452704906463623,
+      "learning_rate": 4.276209357652657e-05,
+      "loss": 1.4,
+      "step": 27391
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9316123723983765,
+      "learning_rate": 4.276182923605604e-05,
+      "loss": 1.3177,
+      "step": 27392
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0618531703948975,
+      "learning_rate": 4.2761564895585515e-05,
+      "loss": 1.3713,
+      "step": 27393
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8040138483047485,
+      "learning_rate": 4.276130055511499e-05,
+      "loss": 1.2433,
+      "step": 27394
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.897321343421936,
+      "learning_rate": 4.276103621464446e-05,
+      "loss": 1.3645,
+      "step": 27395
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9077835083007812,
+      "learning_rate": 4.2760771874173936e-05,
+      "loss": 1.3534,
+      "step": 27396
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4419541358947754,
+      "learning_rate": 4.2760507533703416e-05,
+      "loss": 1.2462,
+      "step": 27397
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3098556995391846,
+      "learning_rate": 4.276024319323288e-05,
+      "loss": 1.3883,
+      "step": 27398
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4756176471710205,
+      "learning_rate": 4.275997885276236e-05,
+      "loss": 1.1941,
+      "step": 27399
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0850794315338135,
+      "learning_rate": 4.275971451229183e-05,
+      "loss": 1.2903,
+      "step": 27400
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.029016017913818,
+      "learning_rate": 4.275945017182131e-05,
+      "loss": 1.5568,
+      "step": 27401
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.549267530441284,
+      "learning_rate": 4.2759185831350784e-05,
+      "loss": 1.3736,
+      "step": 27402
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.367427110671997,
+      "learning_rate": 4.275892149088026e-05,
+      "loss": 1.8215,
+      "step": 27403
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.995335578918457,
+      "learning_rate": 4.275865715040973e-05,
+      "loss": 1.4009,
+      "step": 27404
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.389186382293701,
+      "learning_rate": 4.2758392809939205e-05,
+      "loss": 1.3011,
+      "step": 27405
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.962346076965332,
+      "learning_rate": 4.275812846946868e-05,
+      "loss": 1.277,
+      "step": 27406
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.1434617042541504,
+      "learning_rate": 4.275786412899815e-05,
+      "loss": 1.3857,
+      "step": 27407
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6997723579406738,
+      "learning_rate": 4.2757599788527625e-05,
+      "loss": 1.2173,
+      "step": 27408
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.15287184715271,
+      "learning_rate": 4.27573354480571e-05,
+      "loss": 1.4674,
+      "step": 27409
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2844765186309814,
+      "learning_rate": 4.275707110758657e-05,
+      "loss": 1.2906,
+      "step": 27410
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8416903018951416,
+      "learning_rate": 4.2756806767116046e-05,
+      "loss": 1.3383,
+      "step": 27411
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.5522851943969727,
+      "learning_rate": 4.2756542426645526e-05,
+      "loss": 1.3678,
+      "step": 27412
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9398492574691772,
+      "learning_rate": 4.275627808617499e-05,
+      "loss": 1.3632,
+      "step": 27413
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.156493902206421,
+      "learning_rate": 4.2756013745704474e-05,
+      "loss": 1.4129,
+      "step": 27414
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9898710250854492,
+      "learning_rate": 4.275574940523394e-05,
+      "loss": 1.343,
+      "step": 27415
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0531954765319824,
+      "learning_rate": 4.275548506476342e-05,
+      "loss": 1.1708,
+      "step": 27416
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.726938247680664,
+      "learning_rate": 4.2755220724292894e-05,
+      "loss": 1.5257,
+      "step": 27417
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3301572799682617,
+      "learning_rate": 4.275495638382236e-05,
+      "loss": 1.4529,
+      "step": 27418
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9041792154312134,
+      "learning_rate": 4.275469204335184e-05,
+      "loss": 1.378,
+      "step": 27419
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.599829077720642,
+      "learning_rate": 4.275442770288131e-05,
+      "loss": 1.208,
+      "step": 27420
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9223214387893677,
+      "learning_rate": 4.275416336241079e-05,
+      "loss": 1.2539,
+      "step": 27421
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2966084480285645,
+      "learning_rate": 4.275389902194026e-05,
+      "loss": 2.2759,
+      "step": 27422
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.682995080947876,
+      "learning_rate": 4.2753634681469736e-05,
+      "loss": 1.1996,
+      "step": 27423
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.457026720046997,
+      "learning_rate": 4.275337034099921e-05,
+      "loss": 1.4162,
+      "step": 27424
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8411641120910645,
+      "learning_rate": 4.275310600052868e-05,
+      "loss": 1.3105,
+      "step": 27425
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8860396146774292,
+      "learning_rate": 4.2752841660058156e-05,
+      "loss": 1.1146,
+      "step": 27426
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2069664001464844,
+      "learning_rate": 4.275257731958764e-05,
+      "loss": 1.3523,
+      "step": 27427
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.215360403060913,
+      "learning_rate": 4.2752312979117104e-05,
+      "loss": 1.3131,
+      "step": 27428
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2779173851013184,
+      "learning_rate": 4.275204863864658e-05,
+      "loss": 1.3603,
+      "step": 27429
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9785361289978027,
+      "learning_rate": 4.275178429817605e-05,
+      "loss": 1.3209,
+      "step": 27430
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9623414278030396,
+      "learning_rate": 4.2751519957705524e-05,
+      "loss": 1.1948,
+      "step": 27431
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.1333155632019043,
+      "learning_rate": 4.2751255617235005e-05,
+      "loss": 1.3641,
+      "step": 27432
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.775792360305786,
+      "learning_rate": 4.275099127676447e-05,
+      "loss": 1.4528,
+      "step": 27433
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.922197937965393,
+      "learning_rate": 4.275072693629395e-05,
+      "loss": 1.2533,
+      "step": 27434
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.752657413482666,
+      "learning_rate": 4.275046259582342e-05,
+      "loss": 2.7453,
+      "step": 27435
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8895012140274048,
+      "learning_rate": 4.27501982553529e-05,
+      "loss": 1.214,
+      "step": 27436
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.247812509536743,
+      "learning_rate": 4.274993391488237e-05,
+      "loss": 1.3814,
+      "step": 27437
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.943965196609497,
+      "learning_rate": 4.2749669574411846e-05,
+      "loss": 1.278,
+      "step": 27438
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2601239681243896,
+      "learning_rate": 4.274940523394132e-05,
+      "loss": 1.373,
+      "step": 27439
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.725558042526245,
+      "learning_rate": 4.274914089347079e-05,
+      "loss": 1.4688,
+      "step": 27440
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0806424617767334,
+      "learning_rate": 4.274887655300027e-05,
+      "loss": 1.2859,
+      "step": 27441
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.1334011554718018,
+      "learning_rate": 4.274861221252974e-05,
+      "loss": 1.408,
+      "step": 27442
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.5092674493789673,
+      "learning_rate": 4.2748347872059214e-05,
+      "loss": 1.1765,
+      "step": 27443
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.329010248184204,
+      "learning_rate": 4.274808353158869e-05,
+      "loss": 1.5248,
+      "step": 27444
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9109681844711304,
+      "learning_rate": 4.274781919111816e-05,
+      "loss": 1.2864,
+      "step": 27445
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9150786399841309,
+      "learning_rate": 4.2747554850647635e-05,
+      "loss": 1.3482,
+      "step": 27446
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0394837856292725,
+      "learning_rate": 4.2747290510177115e-05,
+      "loss": 1.2583,
+      "step": 27447
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3042473793029785,
+      "learning_rate": 4.274702616970658e-05,
+      "loss": 1.2119,
+      "step": 27448
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9932929277420044,
+      "learning_rate": 4.274676182923606e-05,
+      "loss": 1.3946,
+      "step": 27449
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.7460453510284424,
+      "learning_rate": 4.274649748876553e-05,
+      "loss": 1.3267,
+      "step": 27450
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9158090353012085,
+      "learning_rate": 4.274623314829501e-05,
+      "loss": 1.1688,
+      "step": 27451
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.589876651763916,
+      "learning_rate": 4.274596880782448e-05,
+      "loss": 1.4301,
+      "step": 27452
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.889370322227478,
+      "learning_rate": 4.2745704467353956e-05,
+      "loss": 1.4137,
+      "step": 27453
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7918031215667725,
+      "learning_rate": 4.274544012688343e-05,
+      "loss": 2.3972,
+      "step": 27454
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8071457147598267,
+      "learning_rate": 4.27451757864129e-05,
+      "loss": 1.4267,
+      "step": 27455
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0846762657165527,
+      "learning_rate": 4.274491144594238e-05,
+      "loss": 1.3282,
+      "step": 27456
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.1693530082702637,
+      "learning_rate": 4.274464710547185e-05,
+      "loss": 1.3234,
+      "step": 27457
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.181340456008911,
+      "learning_rate": 4.2744382765001324e-05,
+      "loss": 1.3339,
+      "step": 27458
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9799538850784302,
+      "learning_rate": 4.27441184245308e-05,
+      "loss": 1.2704,
+      "step": 27459
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0709948539733887,
+      "learning_rate": 4.274385408406027e-05,
+      "loss": 1.2701,
+      "step": 27460
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.5438207387924194,
+      "learning_rate": 4.2743589743589745e-05,
+      "loss": 2.0544,
+      "step": 27461
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6931724548339844,
+      "learning_rate": 4.2743325403119225e-05,
+      "loss": 1.2754,
+      "step": 27462
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9937978982925415,
+      "learning_rate": 4.274306106264869e-05,
+      "loss": 1.2885,
+      "step": 27463
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0873124599456787,
+      "learning_rate": 4.2742796722178166e-05,
+      "loss": 1.2008,
+      "step": 27464
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.2951807975769043,
+      "learning_rate": 4.274253238170764e-05,
+      "loss": 1.411,
+      "step": 27465
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2923617362976074,
+      "learning_rate": 4.274226804123711e-05,
+      "loss": 1.3589,
+      "step": 27466
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.567655563354492,
+      "learning_rate": 4.274200370076659e-05,
+      "loss": 1.5028,
+      "step": 27467
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9591834545135498,
+      "learning_rate": 4.274173936029606e-05,
+      "loss": 1.3131,
+      "step": 27468
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8879563808441162,
+      "learning_rate": 4.274147501982554e-05,
+      "loss": 1.3068,
+      "step": 27469
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9202834367752075,
+      "learning_rate": 4.274121067935501e-05,
+      "loss": 1.2832,
+      "step": 27470
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9382426738739014,
+      "learning_rate": 4.274094633888449e-05,
+      "loss": 1.1649,
+      "step": 27471
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7440166473388672,
+      "learning_rate": 4.274068199841396e-05,
+      "loss": 1.2539,
+      "step": 27472
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.112314224243164,
+      "learning_rate": 4.2740417657943435e-05,
+      "loss": 1.2264,
+      "step": 27473
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2988831996917725,
+      "learning_rate": 4.274015331747291e-05,
+      "loss": 1.9491,
+      "step": 27474
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.54636549949646,
+      "learning_rate": 4.273988897700238e-05,
+      "loss": 1.3493,
+      "step": 27475
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.941322922706604,
+      "learning_rate": 4.2739624636531855e-05,
+      "loss": 1.3038,
+      "step": 27476
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9973632097244263,
+      "learning_rate": 4.273936029606133e-05,
+      "loss": 1.4063,
+      "step": 27477
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6777764558792114,
+      "learning_rate": 4.27390959555908e-05,
+      "loss": 1.7678,
+      "step": 27478
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9543145895004272,
+      "learning_rate": 4.2738831615120276e-05,
+      "loss": 1.3212,
+      "step": 27479
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0084433555603027,
+      "learning_rate": 4.273856727464975e-05,
+      "loss": 1.2233,
+      "step": 27480
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7443740367889404,
+      "learning_rate": 4.273830293417922e-05,
+      "loss": 1.3562,
+      "step": 27481
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6135820150375366,
+      "learning_rate": 4.2738038593708703e-05,
+      "loss": 1.1546,
+      "step": 27482
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.497530460357666,
+      "learning_rate": 4.273777425323817e-05,
+      "loss": 1.4656,
+      "step": 27483
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0893619060516357,
+      "learning_rate": 4.273750991276765e-05,
+      "loss": 1.2617,
+      "step": 27484
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.6958515644073486,
+      "learning_rate": 4.273724557229712e-05,
+      "loss": 1.4008,
+      "step": 27485
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6248366832733154,
+      "learning_rate": 4.27369812318266e-05,
+      "loss": 1.3022,
+      "step": 27486
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8674947023391724,
+      "learning_rate": 4.273671689135607e-05,
+      "loss": 1.3667,
+      "step": 27487
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2567503452301025,
+      "learning_rate": 4.2736452550885545e-05,
+      "loss": 1.3776,
+      "step": 27488
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.6226119995117188,
+      "learning_rate": 4.273618821041502e-05,
+      "loss": 1.4575,
+      "step": 27489
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.603933095932007,
+      "learning_rate": 4.2735923869944485e-05,
+      "loss": 1.3612,
+      "step": 27490
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.156212091445923,
+      "learning_rate": 4.2735659529473966e-05,
+      "loss": 1.4007,
+      "step": 27491
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.772141456604004,
+      "learning_rate": 4.273539518900344e-05,
+      "loss": 1.2801,
+      "step": 27492
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.807969093322754,
+      "learning_rate": 4.273513084853291e-05,
+      "loss": 1.2837,
+      "step": 27493
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.690582036972046,
+      "learning_rate": 4.2734866508062386e-05,
+      "loss": 1.1838,
+      "step": 27494
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8304349184036255,
+      "learning_rate": 4.273460216759186e-05,
+      "loss": 1.1207,
+      "step": 27495
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.9402103424072266,
+      "learning_rate": 4.2734337827121333e-05,
+      "loss": 1.4978,
+      "step": 27496
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.1867847442626953,
+      "learning_rate": 4.2734073486650814e-05,
+      "loss": 1.1644,
+      "step": 27497
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.132800817489624,
+      "learning_rate": 4.273380914618028e-05,
+      "loss": 1.2891,
+      "step": 27498
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.934761881828308,
+      "learning_rate": 4.273354480570976e-05,
+      "loss": 1.2895,
+      "step": 27499
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6093416213989258,
+      "learning_rate": 4.273328046523923e-05,
+      "loss": 1.1672,
+      "step": 27500
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.715700387954712,
+      "learning_rate": 4.27330161247687e-05,
+      "loss": 1.2203,
+      "step": 27501
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9032392501831055,
+      "learning_rate": 4.273275178429818e-05,
+      "loss": 1.3079,
+      "step": 27502
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7930841445922852,
+      "learning_rate": 4.273248744382765e-05,
+      "loss": 1.4171,
+      "step": 27503
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.021359443664551,
+      "learning_rate": 4.273222310335713e-05,
+      "loss": 1.3348,
+      "step": 27504
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8462910652160645,
+      "learning_rate": 4.2731958762886596e-05,
+      "loss": 1.2875,
+      "step": 27505
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.1150918006896973,
+      "learning_rate": 4.2731694422416076e-05,
+      "loss": 1.4083,
+      "step": 27506
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8805415630340576,
+      "learning_rate": 4.273143008194555e-05,
+      "loss": 1.2228,
+      "step": 27507
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.5885210037231445,
+      "learning_rate": 4.273116574147502e-05,
+      "loss": 1.4206,
+      "step": 27508
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0843894481658936,
+      "learning_rate": 4.27309014010045e-05,
+      "loss": 1.2182,
+      "step": 27509
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6799746751785278,
+      "learning_rate": 4.273063706053397e-05,
+      "loss": 1.1952,
+      "step": 27510
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.7677290439605713,
+      "learning_rate": 4.2730372720063444e-05,
+      "loss": 1.6371,
+      "step": 27511
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.6730244159698486,
+      "learning_rate": 4.273010837959292e-05,
+      "loss": 1.3663,
+      "step": 27512
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4674298763275146,
+      "learning_rate": 4.272984403912239e-05,
+      "loss": 1.2951,
+      "step": 27513
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3824193477630615,
+      "learning_rate": 4.2729579698651865e-05,
+      "loss": 1.3791,
+      "step": 27514
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7192330360412598,
+      "learning_rate": 4.272931535818134e-05,
+      "loss": 1.1795,
+      "step": 27515
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.3106467723846436,
+      "learning_rate": 4.272905101771081e-05,
+      "loss": 1.4847,
+      "step": 27516
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0862720012664795,
+      "learning_rate": 4.272878667724029e-05,
+      "loss": 1.2213,
+      "step": 27517
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8245084285736084,
+      "learning_rate": 4.272852233676976e-05,
+      "loss": 1.1285,
+      "step": 27518
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7871942520141602,
+      "learning_rate": 4.272825799629924e-05,
+      "loss": 1.3246,
+      "step": 27519
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.0375537872314453,
+      "learning_rate": 4.2727993655828706e-05,
+      "loss": 1.3124,
+      "step": 27520
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9626736640930176,
+      "learning_rate": 4.2727729315358186e-05,
+      "loss": 1.2031,
+      "step": 27521
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8943884372711182,
+      "learning_rate": 4.272746497488766e-05,
+      "loss": 1.2624,
+      "step": 27522
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3190319538116455,
+      "learning_rate": 4.2727200634417133e-05,
+      "loss": 1.3506,
+      "step": 27523
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.4435243606567383,
+      "learning_rate": 4.272693629394661e-05,
+      "loss": 1.2297,
+      "step": 27524
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.856396198272705,
+      "learning_rate": 4.272667195347608e-05,
+      "loss": 1.3605,
+      "step": 27525
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.091740608215332,
+      "learning_rate": 4.2726407613005554e-05,
+      "loss": 1.3315,
+      "step": 27526
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.756132125854492,
+      "learning_rate": 4.272614327253503e-05,
+      "loss": 1.4547,
+      "step": 27527
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.2521538734436035,
+      "learning_rate": 4.27258789320645e-05,
+      "loss": 1.4117,
+      "step": 27528
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9360243082046509,
+      "learning_rate": 4.2725614591593975e-05,
+      "loss": 1.2376,
+      "step": 27529
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.881056547164917,
+      "learning_rate": 4.272535025112345e-05,
+      "loss": 1.2908,
+      "step": 27530
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7240703105926514,
+      "learning_rate": 4.272508591065292e-05,
+      "loss": 1.1351,
+      "step": 27531
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.930344820022583,
+      "learning_rate": 4.27248215701824e-05,
+      "loss": 1.3592,
+      "step": 27532
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.255746603012085,
+      "learning_rate": 4.272455722971187e-05,
+      "loss": 1.4432,
+      "step": 27533
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.5828979015350342,
+      "learning_rate": 4.272429288924135e-05,
+      "loss": 1.194,
+      "step": 27534
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.5702162981033325,
+      "learning_rate": 4.2724028548770816e-05,
+      "loss": 1.8826,
+      "step": 27535
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7898279428482056,
+      "learning_rate": 4.272376420830029e-05,
+      "loss": 1.2216,
+      "step": 27536
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.962753176689148,
+      "learning_rate": 4.272349986782977e-05,
+      "loss": 1.3564,
+      "step": 27537
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1443028450012207,
+      "learning_rate": 4.272323552735924e-05,
+      "loss": 1.3353,
+      "step": 27538
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.2313239574432373,
+      "learning_rate": 4.272297118688872e-05,
+      "loss": 1.3603,
+      "step": 27539
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8413019180297852,
+      "learning_rate": 4.2722706846418184e-05,
+      "loss": 1.2148,
+      "step": 27540
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.313427209854126,
+      "learning_rate": 4.2722442505947664e-05,
+      "loss": 1.2735,
+      "step": 27541
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.467670202255249,
+      "learning_rate": 4.272217816547714e-05,
+      "loss": 1.4491,
+      "step": 27542
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.070507049560547,
+      "learning_rate": 4.272191382500661e-05,
+      "loss": 1.2651,
+      "step": 27543
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0813019275665283,
+      "learning_rate": 4.2721649484536085e-05,
+      "loss": 1.393,
+      "step": 27544
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8311240673065186,
+      "learning_rate": 4.272138514406556e-05,
+      "loss": 1.2633,
+      "step": 27545
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.635404348373413,
+      "learning_rate": 4.272112080359503e-05,
+      "loss": 1.1779,
+      "step": 27546
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8494809865951538,
+      "learning_rate": 4.2720856463124506e-05,
+      "loss": 1.2114,
+      "step": 27547
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1678972244262695,
+      "learning_rate": 4.272059212265398e-05,
+      "loss": 1.2757,
+      "step": 27548
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6888669729232788,
+      "learning_rate": 4.272032778218345e-05,
+      "loss": 1.2504,
+      "step": 27549
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8879802227020264,
+      "learning_rate": 4.272006344171293e-05,
+      "loss": 1.2986,
+      "step": 27550
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.440401315689087,
+      "learning_rate": 4.27197991012424e-05,
+      "loss": 1.4089,
+      "step": 27551
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9535675048828125,
+      "learning_rate": 4.271953476077188e-05,
+      "loss": 1.2607,
+      "step": 27552
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.4523026943206787,
+      "learning_rate": 4.271927042030135e-05,
+      "loss": 2.556,
+      "step": 27553
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.9060165882110596,
+      "learning_rate": 4.271900607983083e-05,
+      "loss": 1.4311,
+      "step": 27554
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9585230350494385,
+      "learning_rate": 4.2718741739360294e-05,
+      "loss": 1.242,
+      "step": 27555
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9668090343475342,
+      "learning_rate": 4.2718477398889775e-05,
+      "loss": 1.2525,
+      "step": 27556
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.83659827709198,
+      "learning_rate": 4.271821305841925e-05,
+      "loss": 1.3556,
+      "step": 27557
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.371110677719116,
+      "learning_rate": 4.271794871794872e-05,
+      "loss": 1.4475,
+      "step": 27558
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7453049421310425,
+      "learning_rate": 4.2717684377478196e-05,
+      "loss": 1.2335,
+      "step": 27559
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.9505536556243896,
+      "learning_rate": 4.271742003700767e-05,
+      "loss": 1.483,
+      "step": 27560
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.289287567138672,
+      "learning_rate": 4.271715569653714e-05,
+      "loss": 1.2493,
+      "step": 27561
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.728266716003418,
+      "learning_rate": 4.2716891356066616e-05,
+      "loss": 1.2345,
+      "step": 27562
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8748241662979126,
+      "learning_rate": 4.271662701559609e-05,
+      "loss": 1.3324,
+      "step": 27563
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.244598150253296,
+      "learning_rate": 4.271636267512556e-05,
+      "loss": 1.4014,
+      "step": 27564
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.694384217262268,
+      "learning_rate": 4.271609833465504e-05,
+      "loss": 1.1637,
+      "step": 27565
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9003735780715942,
+      "learning_rate": 4.271583399418451e-05,
+      "loss": 1.0934,
+      "step": 27566
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.375037670135498,
+      "learning_rate": 4.271556965371399e-05,
+      "loss": 1.4639,
+      "step": 27567
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0324900150299072,
+      "learning_rate": 4.271530531324346e-05,
+      "loss": 1.3359,
+      "step": 27568
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.681043267250061,
+      "learning_rate": 4.271504097277294e-05,
+      "loss": 1.3568,
+      "step": 27569
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.5290915966033936,
+      "learning_rate": 4.2714776632302405e-05,
+      "loss": 1.0711,
+      "step": 27570
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9666457176208496,
+      "learning_rate": 4.2714512291831885e-05,
+      "loss": 1.4319,
+      "step": 27571
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6819512844085693,
+      "learning_rate": 4.271424795136136e-05,
+      "loss": 1.1506,
+      "step": 27572
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.5851967334747314,
+      "learning_rate": 4.2713983610890826e-05,
+      "loss": 1.4298,
+      "step": 27573
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.3908212184906006,
+      "learning_rate": 4.2713719270420306e-05,
+      "loss": 1.2595,
+      "step": 27574
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6986539363861084,
+      "learning_rate": 4.271345492994977e-05,
+      "loss": 1.2305,
+      "step": 27575
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8448537588119507,
+      "learning_rate": 4.271319058947925e-05,
+      "loss": 1.2888,
+      "step": 27576
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9185301065444946,
+      "learning_rate": 4.2712926249008727e-05,
+      "loss": 1.2608,
+      "step": 27577
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.5975228548049927,
+      "learning_rate": 4.27126619085382e-05,
+      "loss": 1.2592,
+      "step": 27578
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.2766709327697754,
+      "learning_rate": 4.2712397568067674e-05,
+      "loss": 2.2803,
+      "step": 27579
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1200075149536133,
+      "learning_rate": 4.271213322759715e-05,
+      "loss": 1.5261,
+      "step": 27580
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8949902057647705,
+      "learning_rate": 4.271186888712662e-05,
+      "loss": 1.2048,
+      "step": 27581
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.11610746383667,
+      "learning_rate": 4.2711604546656094e-05,
+      "loss": 1.3765,
+      "step": 27582
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.1395301818847656,
+      "learning_rate": 4.271134020618557e-05,
+      "loss": 1.4339,
+      "step": 27583
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9811193943023682,
+      "learning_rate": 4.271107586571504e-05,
+      "loss": 1.189,
+      "step": 27584
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.9178717136383057,
+      "learning_rate": 4.2710811525244515e-05,
+      "loss": 1.6089,
+      "step": 27585
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7300492525100708,
+      "learning_rate": 4.271054718477399e-05,
+      "loss": 1.1825,
+      "step": 27586
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0752291679382324,
+      "learning_rate": 4.271028284430347e-05,
+      "loss": 1.3312,
+      "step": 27587
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.649970293045044,
+      "learning_rate": 4.2710018503832936e-05,
+      "loss": 1.3334,
+      "step": 27588
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.259871244430542,
+      "learning_rate": 4.2709754163362416e-05,
+      "loss": 1.1996,
+      "step": 27589
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0539207458496094,
+      "learning_rate": 4.270948982289188e-05,
+      "loss": 1.3565,
+      "step": 27590
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.2020423412323,
+      "learning_rate": 4.270922548242136e-05,
+      "loss": 1.3199,
+      "step": 27591
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.4181067943573,
+      "learning_rate": 4.270896114195084e-05,
+      "loss": 1.4324,
+      "step": 27592
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.9337148666381836,
+      "learning_rate": 4.270869680148031e-05,
+      "loss": 1.4096,
+      "step": 27593
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.3220055103302,
+      "learning_rate": 4.2708432461009784e-05,
+      "loss": 1.3717,
+      "step": 27594
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.057434558868408,
+      "learning_rate": 4.270816812053926e-05,
+      "loss": 1.4119,
+      "step": 27595
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.836167097091675,
+      "learning_rate": 4.270790378006873e-05,
+      "loss": 1.4086,
+      "step": 27596
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.615828037261963,
+      "learning_rate": 4.2707639439598205e-05,
+      "loss": 1.3483,
+      "step": 27597
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.14471435546875,
+      "learning_rate": 4.270737509912768e-05,
+      "loss": 1.2965,
+      "step": 27598
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.2977592945098877,
+      "learning_rate": 4.270711075865715e-05,
+      "loss": 1.5592,
+      "step": 27599
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.2994225025177,
+      "learning_rate": 4.2706846418186625e-05,
+      "loss": 1.3839,
+      "step": 27600
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7661830186843872,
+      "learning_rate": 4.27065820777161e-05,
+      "loss": 1.1966,
+      "step": 27601
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8131206035614014,
+      "learning_rate": 4.270631773724558e-05,
+      "loss": 1.267,
+      "step": 27602
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.6660690307617188,
+      "learning_rate": 4.2706053396775046e-05,
+      "loss": 1.2398,
+      "step": 27603
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.5390095710754395,
+      "learning_rate": 4.2705789056304527e-05,
+      "loss": 1.6648,
+      "step": 27604
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9341356754302979,
+      "learning_rate": 4.270552471583399e-05,
+      "loss": 1.3729,
+      "step": 27605
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0546467304229736,
+      "learning_rate": 4.2705260375363474e-05,
+      "loss": 1.2781,
+      "step": 27606
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0243096351623535,
+      "learning_rate": 4.270499603489295e-05,
+      "loss": 1.2835,
+      "step": 27607
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6911801099777222,
+      "learning_rate": 4.2704731694422414e-05,
+      "loss": 1.2769,
+      "step": 27608
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.8930275440216064,
+      "learning_rate": 4.2704467353951894e-05,
+      "loss": 1.4077,
+      "step": 27609
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.3195958137512207,
+      "learning_rate": 4.270420301348136e-05,
+      "loss": 1.4636,
+      "step": 27610
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.910512089729309,
+      "learning_rate": 4.270393867301084e-05,
+      "loss": 1.3432,
+      "step": 27611
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1201295852661133,
+      "learning_rate": 4.2703674332540315e-05,
+      "loss": 1.2983,
+      "step": 27612
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.039560317993164,
+      "learning_rate": 4.270340999206979e-05,
+      "loss": 1.4964,
+      "step": 27613
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.4388794898986816,
+      "learning_rate": 4.270314565159926e-05,
+      "loss": 1.3651,
+      "step": 27614
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.408916711807251,
+      "learning_rate": 4.2702881311128736e-05,
+      "loss": 1.4245,
+      "step": 27615
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.4277865886688232,
+      "learning_rate": 4.270261697065821e-05,
+      "loss": 1.4459,
+      "step": 27616
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.3849494457244873,
+      "learning_rate": 4.270235263018769e-05,
+      "loss": 1.3904,
+      "step": 27617
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.904853343963623,
+      "learning_rate": 4.2702088289717157e-05,
+      "loss": 1.4702,
+      "step": 27618
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8827029466629028,
+      "learning_rate": 4.270182394924663e-05,
+      "loss": 1.2942,
+      "step": 27619
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0005431175231934,
+      "learning_rate": 4.2701559608776104e-05,
+      "loss": 1.2218,
+      "step": 27620
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6810225248336792,
+      "learning_rate": 4.270129526830558e-05,
+      "loss": 1.1848,
+      "step": 27621
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.082423448562622,
+      "learning_rate": 4.270103092783506e-05,
+      "loss": 1.3993,
+      "step": 27622
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.653648614883423,
+      "learning_rate": 4.2700766587364524e-05,
+      "loss": 1.4692,
+      "step": 27623
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.778665542602539,
+      "learning_rate": 4.2700502246894005e-05,
+      "loss": 1.1515,
+      "step": 27624
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8413453102111816,
+      "learning_rate": 4.270023790642347e-05,
+      "loss": 1.4078,
+      "step": 27625
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.3145337104797363,
+      "learning_rate": 4.269997356595295e-05,
+      "loss": 1.4346,
+      "step": 27626
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8130873441696167,
+      "learning_rate": 4.2699709225482425e-05,
+      "loss": 1.2943,
+      "step": 27627
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1357421875,
+      "learning_rate": 4.26994448850119e-05,
+      "loss": 1.3214,
+      "step": 27628
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.413775682449341,
+      "learning_rate": 4.269918054454137e-05,
+      "loss": 1.3027,
+      "step": 27629
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7088183164596558,
+      "learning_rate": 4.2698916204070846e-05,
+      "loss": 1.2123,
+      "step": 27630
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.371962785720825,
+      "learning_rate": 4.269865186360032e-05,
+      "loss": 1.4713,
+      "step": 27631
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1349260807037354,
+      "learning_rate": 4.269838752312979e-05,
+      "loss": 1.3656,
+      "step": 27632
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8430343866348267,
+      "learning_rate": 4.269812318265927e-05,
+      "loss": 1.2649,
+      "step": 27633
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.5718226432800293,
+      "learning_rate": 4.269785884218874e-05,
+      "loss": 1.2227,
+      "step": 27634
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.7798075675964355,
+      "learning_rate": 4.2697594501718214e-05,
+      "loss": 1.5011,
+      "step": 27635
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.857538938522339,
+      "learning_rate": 4.269733016124769e-05,
+      "loss": 1.6886,
+      "step": 27636
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.2210099697113037,
+      "learning_rate": 4.269706582077717e-05,
+      "loss": 1.3837,
+      "step": 27637
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.3610830307006836,
+      "learning_rate": 4.2696801480306635e-05,
+      "loss": 1.3897,
+      "step": 27638
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8274344205856323,
+      "learning_rate": 4.2696537139836115e-05,
+      "loss": 1.2321,
+      "step": 27639
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8806899785995483,
+      "learning_rate": 4.269627279936558e-05,
+      "loss": 1.3309,
+      "step": 27640
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.858370065689087,
+      "learning_rate": 4.269600845889506e-05,
+      "loss": 1.4564,
+      "step": 27641
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7406306266784668,
+      "learning_rate": 4.2695744118424536e-05,
+      "loss": 1.0488,
+      "step": 27642
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.475791096687317,
+      "learning_rate": 4.269547977795401e-05,
+      "loss": 1.7069,
+      "step": 27643
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0423362255096436,
+      "learning_rate": 4.269521543748348e-05,
+      "loss": 1.1964,
+      "step": 27644
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.404784917831421,
+      "learning_rate": 4.269495109701295e-05,
+      "loss": 1.211,
+      "step": 27645
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.002239227294922,
+      "learning_rate": 4.269468675654243e-05,
+      "loss": 1.3082,
+      "step": 27646
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.454336404800415,
+      "learning_rate": 4.2694422416071904e-05,
+      "loss": 1.3215,
+      "step": 27647
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.5469703674316406,
+      "learning_rate": 4.269415807560138e-05,
+      "loss": 1.3371,
+      "step": 27648
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.6013665199279785,
+      "learning_rate": 4.269389373513085e-05,
+      "loss": 1.427,
+      "step": 27649
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.165423631668091,
+      "learning_rate": 4.2693629394660324e-05,
+      "loss": 1.3919,
+      "step": 27650
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1040210723876953,
+      "learning_rate": 4.26933650541898e-05,
+      "loss": 1.4166,
+      "step": 27651
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.178431749343872,
+      "learning_rate": 4.269310071371928e-05,
+      "loss": 1.3639,
+      "step": 27652
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9357876777648926,
+      "learning_rate": 4.2692836373248745e-05,
+      "loss": 1.2187,
+      "step": 27653
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.605695366859436,
+      "learning_rate": 4.269257203277822e-05,
+      "loss": 1.1885,
+      "step": 27654
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6989729404449463,
+      "learning_rate": 4.269230769230769e-05,
+      "loss": 1.2712,
+      "step": 27655
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.587153434753418,
+      "learning_rate": 4.2692043351837166e-05,
+      "loss": 1.4817,
+      "step": 27656
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1784746646881104,
+      "learning_rate": 4.2691779011366646e-05,
+      "loss": 1.2673,
+      "step": 27657
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0734448432922363,
+      "learning_rate": 4.269151467089611e-05,
+      "loss": 1.2917,
+      "step": 27658
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8938947916030884,
+      "learning_rate": 4.269125033042559e-05,
+      "loss": 1.2692,
+      "step": 27659
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1634042263031006,
+      "learning_rate": 4.269098598995506e-05,
+      "loss": 1.3465,
+      "step": 27660
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7119481563568115,
+      "learning_rate": 4.269072164948454e-05,
+      "loss": 1.1839,
+      "step": 27661
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.4340693950653076,
+      "learning_rate": 4.2690457309014014e-05,
+      "loss": 1.4105,
+      "step": 27662
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6363277435302734,
+      "learning_rate": 4.269019296854349e-05,
+      "loss": 1.1398,
+      "step": 27663
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.5613510608673096,
+      "learning_rate": 4.268992862807296e-05,
+      "loss": 1.3276,
+      "step": 27664
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.3110063076019287,
+      "learning_rate": 4.2689664287602435e-05,
+      "loss": 1.395,
+      "step": 27665
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.032055139541626,
+      "learning_rate": 4.268939994713191e-05,
+      "loss": 1.2903,
+      "step": 27666
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9142987728118896,
+      "learning_rate": 4.268913560666138e-05,
+      "loss": 1.2783,
+      "step": 27667
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8759918212890625,
+      "learning_rate": 4.2688871266190855e-05,
+      "loss": 1.3913,
+      "step": 27668
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0188803672790527,
+      "learning_rate": 4.268860692572033e-05,
+      "loss": 1.3318,
+      "step": 27669
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.560329914093018,
+      "learning_rate": 4.26883425852498e-05,
+      "loss": 1.7082,
+      "step": 27670
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6794090270996094,
+      "learning_rate": 4.2688078244779276e-05,
+      "loss": 1.2698,
+      "step": 27671
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.7451674938201904,
+      "learning_rate": 4.2687813904308756e-05,
+      "loss": 1.403,
+      "step": 27672
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4715007543563843,
+      "learning_rate": 4.268754956383822e-05,
+      "loss": 1.1968,
+      "step": 27673
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.503421664237976,
+      "learning_rate": 4.2687285223367704e-05,
+      "loss": 1.1809,
+      "step": 27674
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.361030101776123,
+      "learning_rate": 4.268702088289717e-05,
+      "loss": 1.2603,
+      "step": 27675
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8571864366531372,
+      "learning_rate": 4.268675654242665e-05,
+      "loss": 1.274,
+      "step": 27676
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0734565258026123,
+      "learning_rate": 4.2686492201956124e-05,
+      "loss": 1.4557,
+      "step": 27677
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0283124446868896,
+      "learning_rate": 4.26862278614856e-05,
+      "loss": 1.3155,
+      "step": 27678
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.648892641067505,
+      "learning_rate": 4.268596352101507e-05,
+      "loss": 1.2874,
+      "step": 27679
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.273587703704834,
+      "learning_rate": 4.268569918054454e-05,
+      "loss": 1.2113,
+      "step": 27680
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8894343376159668,
+      "learning_rate": 4.268543484007402e-05,
+      "loss": 1.3793,
+      "step": 27681
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0942482948303223,
+      "learning_rate": 4.268517049960349e-05,
+      "loss": 1.2878,
+      "step": 27682
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0948870182037354,
+      "learning_rate": 4.2684906159132966e-05,
+      "loss": 1.2249,
+      "step": 27683
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.252354383468628,
+      "learning_rate": 4.268464181866244e-05,
+      "loss": 1.3155,
+      "step": 27684
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.240182876586914,
+      "learning_rate": 4.268437747819191e-05,
+      "loss": 1.1075,
+      "step": 27685
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.836412787437439,
+      "learning_rate": 4.2684113137721386e-05,
+      "loss": 1.3707,
+      "step": 27686
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6312757730484009,
+      "learning_rate": 4.268384879725087e-05,
+      "loss": 1.109,
+      "step": 27687
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4975298643112183,
+      "learning_rate": 4.2683584456780334e-05,
+      "loss": 1.1361,
+      "step": 27688
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.609012484550476,
+      "learning_rate": 4.2683320116309814e-05,
+      "loss": 1.3045,
+      "step": 27689
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.3074629306793213,
+      "learning_rate": 4.268305577583928e-05,
+      "loss": 1.275,
+      "step": 27690
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9416227340698242,
+      "learning_rate": 4.2682791435368754e-05,
+      "loss": 1.3192,
+      "step": 27691
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9317388534545898,
+      "learning_rate": 4.2682527094898235e-05,
+      "loss": 1.2282,
+      "step": 27692
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.807531714439392,
+      "learning_rate": 4.26822627544277e-05,
+      "loss": 1.4132,
+      "step": 27693
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.483086585998535,
+      "learning_rate": 4.268199841395718e-05,
+      "loss": 1.3916,
+      "step": 27694
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.964056372642517,
+      "learning_rate": 4.268173407348665e-05,
+      "loss": 1.1974,
+      "step": 27695
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0103747844696045,
+      "learning_rate": 4.268146973301613e-05,
+      "loss": 2.1532,
+      "step": 27696
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9877442121505737,
+      "learning_rate": 4.26812053925456e-05,
+      "loss": 1.3583,
+      "step": 27697
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0594370365142822,
+      "learning_rate": 4.2680941052075076e-05,
+      "loss": 1.3235,
+      "step": 27698
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.921936273574829,
+      "learning_rate": 4.268067671160455e-05,
+      "loss": 1.2928,
+      "step": 27699
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8842072486877441,
+      "learning_rate": 4.268041237113402e-05,
+      "loss": 1.195,
+      "step": 27700
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.976211428642273,
+      "learning_rate": 4.26801480306635e-05,
+      "loss": 1.2006,
+      "step": 27701
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.4614057540893555,
+      "learning_rate": 4.267988369019297e-05,
+      "loss": 1.3203,
+      "step": 27702
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.685695171356201,
+      "learning_rate": 4.2679619349722444e-05,
+      "loss": 1.3866,
+      "step": 27703
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.6972744464874268,
+      "learning_rate": 4.267935500925192e-05,
+      "loss": 1.3781,
+      "step": 27704
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1490466594696045,
+      "learning_rate": 4.267909066878139e-05,
+      "loss": 1.3531,
+      "step": 27705
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.2588629722595215,
+      "learning_rate": 4.2678826328310865e-05,
+      "loss": 1.3801,
+      "step": 27706
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1466386318206787,
+      "learning_rate": 4.2678561987840345e-05,
+      "loss": 1.2636,
+      "step": 27707
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.2657229900360107,
+      "learning_rate": 4.267829764736981e-05,
+      "loss": 1.3821,
+      "step": 27708
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7572025060653687,
+      "learning_rate": 4.267803330689929e-05,
+      "loss": 1.1913,
+      "step": 27709
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.242579698562622,
+      "learning_rate": 4.267776896642876e-05,
+      "loss": 1.3325,
+      "step": 27710
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1890077590942383,
+      "learning_rate": 4.267750462595824e-05,
+      "loss": 1.2303,
+      "step": 27711
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.163811445236206,
+      "learning_rate": 4.267724028548771e-05,
+      "loss": 1.4212,
+      "step": 27712
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.1757359504699707,
+      "learning_rate": 4.2676975945017186e-05,
+      "loss": 1.3498,
+      "step": 27713
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.6069576740264893,
+      "learning_rate": 4.267671160454666e-05,
+      "loss": 1.5412,
+      "step": 27714
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.777089238166809,
+      "learning_rate": 4.2676447264076134e-05,
+      "loss": 1.2728,
+      "step": 27715
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.2630116939544678,
+      "learning_rate": 4.267618292360561e-05,
+      "loss": 1.3994,
+      "step": 27716
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.134330987930298,
+      "learning_rate": 4.267591858313508e-05,
+      "loss": 1.2991,
+      "step": 27717
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8842337131500244,
+      "learning_rate": 4.2675654242664554e-05,
+      "loss": 1.2283,
+      "step": 27718
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.1146931648254395,
+      "learning_rate": 4.267538990219403e-05,
+      "loss": 1.3928,
+      "step": 27719
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.4255340099334717,
+      "learning_rate": 4.26751255617235e-05,
+      "loss": 1.1053,
+      "step": 27720
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.5475517511367798,
+      "learning_rate": 4.2674861221252975e-05,
+      "loss": 1.169,
+      "step": 27721
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.1632025241851807,
+      "learning_rate": 4.2674596880782455e-05,
+      "loss": 1.3146,
+      "step": 27722
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.510710597038269,
+      "learning_rate": 4.267433254031192e-05,
+      "loss": 1.148,
+      "step": 27723
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.2853970527648926,
+      "learning_rate": 4.26740681998414e-05,
+      "loss": 1.3399,
+      "step": 27724
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.6633299589157104,
+      "learning_rate": 4.267380385937087e-05,
+      "loss": 1.0659,
+      "step": 27725
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.2298433780670166,
+      "learning_rate": 4.267353951890034e-05,
+      "loss": 1.2611,
+      "step": 27726
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.1502439975738525,
+      "learning_rate": 4.267327517842982e-05,
+      "loss": 1.4592,
+      "step": 27727
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9788668155670166,
+      "learning_rate": 4.267301083795929e-05,
+      "loss": 1.3531,
+      "step": 27728
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.049143075942993,
+      "learning_rate": 4.267274649748877e-05,
+      "loss": 1.3948,
+      "step": 27729
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.6756958961486816,
+      "learning_rate": 4.267248215701824e-05,
+      "loss": 1.3326,
+      "step": 27730
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8198485374450684,
+      "learning_rate": 4.267221781654772e-05,
+      "loss": 1.2725,
+      "step": 27731
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.032465934753418,
+      "learning_rate": 4.267195347607719e-05,
+      "loss": 2.0759,
+      "step": 27732
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.1889102458953857,
+      "learning_rate": 4.2671689135606665e-05,
+      "loss": 1.1248,
+      "step": 27733
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.5968068838119507,
+      "learning_rate": 4.267142479513614e-05,
+      "loss": 1.1528,
+      "step": 27734
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.4103678464889526,
+      "learning_rate": 4.267116045466561e-05,
+      "loss": 1.1266,
+      "step": 27735
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.073535203933716,
+      "learning_rate": 4.2670896114195085e-05,
+      "loss": 1.2297,
+      "step": 27736
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9149292707443237,
+      "learning_rate": 4.267063177372456e-05,
+      "loss": 1.1172,
+      "step": 27737
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.973068356513977,
+      "learning_rate": 4.267036743325403e-05,
+      "loss": 1.2731,
+      "step": 27738
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9631259441375732,
+      "learning_rate": 4.2670103092783506e-05,
+      "loss": 1.8768,
+      "step": 27739
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.152956485748291,
+      "learning_rate": 4.266983875231298e-05,
+      "loss": 1.2772,
+      "step": 27740
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.358506202697754,
+      "learning_rate": 4.266957441184245e-05,
+      "loss": 1.2412,
+      "step": 27741
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.6788980960845947,
+      "learning_rate": 4.2669310071371934e-05,
+      "loss": 1.2434,
+      "step": 27742
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.99558687210083,
+      "learning_rate": 4.26690457309014e-05,
+      "loss": 1.2646,
+      "step": 27743
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.9437484741210938,
+      "learning_rate": 4.266878139043088e-05,
+      "loss": 1.5414,
+      "step": 27744
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.802698850631714,
+      "learning_rate": 4.266851704996035e-05,
+      "loss": 1.4124,
+      "step": 27745
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.2960264682769775,
+      "learning_rate": 4.266825270948983e-05,
+      "loss": 1.233,
+      "step": 27746
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.091961622238159,
+      "learning_rate": 4.26679883690193e-05,
+      "loss": 1.2765,
+      "step": 27747
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7848637104034424,
+      "learning_rate": 4.2667724028548775e-05,
+      "loss": 1.1932,
+      "step": 27748
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8908370733261108,
+      "learning_rate": 4.266745968807825e-05,
+      "loss": 1.163,
+      "step": 27749
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.3653860092163086,
+      "learning_rate": 4.266719534760772e-05,
+      "loss": 1.4399,
+      "step": 27750
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.1247005462646484,
+      "learning_rate": 4.2666931007137196e-05,
+      "loss": 1.2951,
+      "step": 27751
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8463603258132935,
+      "learning_rate": 4.266666666666667e-05,
+      "loss": 1.2107,
+      "step": 27752
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.6322600841522217,
+      "learning_rate": 4.266640232619614e-05,
+      "loss": 1.3577,
+      "step": 27753
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.164094924926758,
+      "learning_rate": 4.2666137985725616e-05,
+      "loss": 1.1973,
+      "step": 27754
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.6806421279907227,
+      "learning_rate": 4.266587364525509e-05,
+      "loss": 1.2773,
+      "step": 27755
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8994404077529907,
+      "learning_rate": 4.2665609304784564e-05,
+      "loss": 1.1617,
+      "step": 27756
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.2690956592559814,
+      "learning_rate": 4.2665344964314044e-05,
+      "loss": 1.2617,
+      "step": 27757
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.912752628326416,
+      "learning_rate": 4.266508062384351e-05,
+      "loss": 1.3455,
+      "step": 27758
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9334105253219604,
+      "learning_rate": 4.266481628337299e-05,
+      "loss": 1.3095,
+      "step": 27759
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.1222329139709473,
+      "learning_rate": 4.266455194290246e-05,
+      "loss": 1.3122,
+      "step": 27760
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.300112009048462,
+      "learning_rate": 4.266428760243194e-05,
+      "loss": 1.6367,
+      "step": 27761
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9811034202575684,
+      "learning_rate": 4.266402326196141e-05,
+      "loss": 1.2865,
+      "step": 27762
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.28922963142395,
+      "learning_rate": 4.266375892149088e-05,
+      "loss": 1.4323,
+      "step": 27763
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.050718069076538,
+      "learning_rate": 4.266349458102036e-05,
+      "loss": 1.3868,
+      "step": 27764
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9389022588729858,
+      "learning_rate": 4.2663230240549826e-05,
+      "loss": 1.3228,
+      "step": 27765
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.460904121398926,
+      "learning_rate": 4.2662965900079306e-05,
+      "loss": 1.3404,
+      "step": 27766
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.165756940841675,
+      "learning_rate": 4.266270155960878e-05,
+      "loss": 1.2553,
+      "step": 27767
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.130305767059326,
+      "learning_rate": 4.266243721913825e-05,
+      "loss": 1.2433,
+      "step": 27768
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9728450775146484,
+      "learning_rate": 4.266217287866773e-05,
+      "loss": 1.3376,
+      "step": 27769
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.090484619140625,
+      "learning_rate": 4.26619085381972e-05,
+      "loss": 1.3762,
+      "step": 27770
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9987449645996094,
+      "learning_rate": 4.2661644197726674e-05,
+      "loss": 1.2637,
+      "step": 27771
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.28179931640625,
+      "learning_rate": 4.266137985725615e-05,
+      "loss": 1.5555,
+      "step": 27772
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.5316901206970215,
+      "learning_rate": 4.266111551678562e-05,
+      "loss": 1.2357,
+      "step": 27773
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.6023377180099487,
+      "learning_rate": 4.2660851176315095e-05,
+      "loss": 1.3175,
+      "step": 27774
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.583254337310791,
+      "learning_rate": 4.266058683584457e-05,
+      "loss": 1.5547,
+      "step": 27775
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.214433431625366,
+      "learning_rate": 4.266032249537404e-05,
+      "loss": 1.3047,
+      "step": 27776
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.230180025100708,
+      "learning_rate": 4.266005815490352e-05,
+      "loss": 1.8942,
+      "step": 27777
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.1349494457244873,
+      "learning_rate": 4.265979381443299e-05,
+      "loss": 1.2501,
+      "step": 27778
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.0522267818450928,
+      "learning_rate": 4.265952947396247e-05,
+      "loss": 1.3833,
+      "step": 27779
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7836321592330933,
+      "learning_rate": 4.2659265133491936e-05,
+      "loss": 1.2073,
+      "step": 27780
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9167869091033936,
+      "learning_rate": 4.2659000793021416e-05,
+      "loss": 1.3129,
+      "step": 27781
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.8376059532165527,
+      "learning_rate": 4.265873645255089e-05,
+      "loss": 1.3472,
+      "step": 27782
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.143842935562134,
+      "learning_rate": 4.2658472112080364e-05,
+      "loss": 1.3718,
+      "step": 27783
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.910766363143921,
+      "learning_rate": 4.265820777160984e-05,
+      "loss": 1.2343,
+      "step": 27784
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8344991207122803,
+      "learning_rate": 4.265794343113931e-05,
+      "loss": 1.3586,
+      "step": 27785
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.9907772541046143,
+      "learning_rate": 4.2657679090668784e-05,
+      "loss": 1.4079,
+      "step": 27786
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9860377311706543,
+      "learning_rate": 4.265741475019826e-05,
+      "loss": 1.3765,
+      "step": 27787
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.2909328937530518,
+      "learning_rate": 4.265715040972773e-05,
+      "loss": 1.4485,
+      "step": 27788
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9208219051361084,
+      "learning_rate": 4.2656886069257205e-05,
+      "loss": 1.3568,
+      "step": 27789
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.917630910873413,
+      "learning_rate": 4.265662172878668e-05,
+      "loss": 1.281,
+      "step": 27790
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.183851480484009,
+      "learning_rate": 4.265635738831615e-05,
+      "loss": 1.2,
+      "step": 27791
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.263993978500366,
+      "learning_rate": 4.2656093047845626e-05,
+      "loss": 1.4457,
+      "step": 27792
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.00960111618042,
+      "learning_rate": 4.26558287073751e-05,
+      "loss": 1.4182,
+      "step": 27793
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.3759889602661133,
+      "learning_rate": 4.265556436690458e-05,
+      "loss": 1.4043,
+      "step": 27794
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9018480777740479,
+      "learning_rate": 4.2655300026434046e-05,
+      "loss": 1.3353,
+      "step": 27795
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.3580799102783203,
+      "learning_rate": 4.265503568596353e-05,
+      "loss": 1.3812,
+      "step": 27796
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.0923266410827637,
+      "learning_rate": 4.2654771345492993e-05,
+      "loss": 1.2653,
+      "step": 27797
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.6045308113098145,
+      "learning_rate": 4.265450700502247e-05,
+      "loss": 1.3345,
+      "step": 27798
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.6155608892440796,
+      "learning_rate": 4.265424266455195e-05,
+      "loss": 1.162,
+      "step": 27799
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9618321657180786,
+      "learning_rate": 4.2653978324081414e-05,
+      "loss": 1.2042,
+      "step": 27800
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 189160,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 50,
+  "total_flos": 2.671864114005639e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}